Searching for a Listed Infrastructure Asset Class: Mean-variance ...

A Publication of the EDHEC Infrastructure Institute-Singapore

Searching for a ListedInfrastructure Asset Class

Mean-variance spanning tests of 22 listed infrastructure proxies

June 2016

infraSingapore Infrastructure Investment Institute

The author would like to thank Noel Amenc, Lionel Martellini, Timo Välilä and two anonymous reviewers for useful comments and suggestions.This study presents the author’s views and conclusions which are not necessarily those of EDHEC Business School.

Table of Contents

Executive Summary . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 4

1 Introduction . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 11

2 Literature Review . . . . . . . . . . . . . . . . . . . . . . . . . . . 14

3 Approach . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 18

4 Methodology . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 21

5 Data . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 24

6 Results . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 34

7 Conclusions . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 50

References . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 61

About the EDHEC Infrastructure Institute-Singapore . . . . . . . . . . 63

Infrastructure Research Publications at EDHEC . . . . . . . . . . . . . 68

Searching for a listed infrastructure asset class - June 2016

About the Authors

Frédéric Blanc-Brude is Director of the EDHEC Infrastructure

Institute–Singapore and represents EDHEC Business School on the

Advisory Board of the Global Infrastructure Facility of the World Bank.

He holds a PhD in Finance (King’s College London), and degrees from

London School of Economics, the Sorbonne and Sciences Po Paris.

Tim Whittaker is an Associate Research Director at EDHEC Infras-

tructure Institute-Singapore and Head of Data Collection. He holds a

Master of Business (Financial Management) an, Bachelors of Economics

and Commerce from the University of Queensland, and a PhD in Finance

from Griffith University.

Simon Wilde is a Ph.D. candidate at the University of Bath, UK and a

Senior Managing Director at Macquarie Capital in London. He is also

part-time Senior Lecturer at UWE Bristol Business School. He holds

degrees from the London School of Economics and the Universities of

Cambridge and Bristol.

A Publication of the EDHEC Infrastructure Institute-Singapore 3

Executive Summary

4 A Publication of the EDHEC Infrastructure Institute-Singapore


Executive Summary

In this paper, we ask the question: does

focusing on listed infrastructure stockscreate diversification benefits previously

unavailable to large investors already active

in public markets?

This question arises from what we call

the ”infrastructure investment narrative”

(Blanc-Brude, 2013), a set of investment

beliefs commonly held by investors about

the investment characteristics of infras-

tructure assets.

In this narrative, the ”infrastructure asset

class” is less exposed to the business cycle

because of the low price-elasticity of infras-

tructure services. Furthermore, the value of

these investment is expected to be mostly

determined by income streams extending

far into the future, and should thus be less

impacted by current events.

According to this intuition, listed infras-tructure may provide diversification

benefits to investors since they are

expected to exhibit low return covariance

with other financial assets. In other words,

listed infrastructure is expected to exhibit

sufficiently unique characteristics to be

considered an ”asset class” in its own right.

Empirically, there are at least three reasons

why this view requires further examination:

1. Most existing research on infrastructure

has used public equity markets to infer

findings for the whole infrastructure

investment universe, but robust and

conclusive evidence is not forthcoming

in existing papers;

2. Index providers have created dedicated

indices focusing on this theme and a

number of active managers propose to

invest in ”listed infrastructure” arguing

that it does indeed constitute a unique

asset class;

3. Listed infrastructure stocks are often

used by investors to proxy investments

in privately held (unlisted) infrastructure

equity, but the adequacy of such proxies

remains untested.

The existence of a distinctive listed infras-tructure effect in investors’ portfolio would

support these views. In the negative, if

this effect cannot be found, there is little

to expect from listed infrastructure equity

from an asset allocation (risk/reward optimi-

sation) perspective and maybe even less

to learn from public markets about the

expected performance of unlisted infras-

tructure investments.

Testing 22 proxies of listedinfrastructureWe test the impact of adding 22 different

proxies of ”listed infrastructure” to the

portfolio of a well-diversified investor using

mean-variance spanning tests. We focus on

three definitions of ”listed infrastructure” as

an asset selection scheme:

1. A ”naïve”, rule-based filtering of stocks

based on industrial sector classifications

and percentage income generated from


Searching for a listed infrastructure asset class- June 2016

Executive Summary

pre-defined infrastructure sectors (nine

proxies);

2. Existing listed infrastructure indices

designed and maintained by index

providers (twelve proxies);

3. A basket of stocks offering a pure

exposure to several hundred underlying

projects that correspond to a well-

known form of infrastructure investment

defined – in contrast with the two

previous cases – in terms of long-term

public-private contracts, not industrial

sectors (one proxy).

Employing the mean-variance spanning

tests originally described by Huberman and

Kandel (1987) and Kan and Zhou (2012),

we test the diversification benefits of these

proxies of the listed infrastructure effect.

There is no listed infrastructureasset classStylised findings include:

1. Our 22 tests of listed infrastructure

reveal little to no robust evidence of

a ”listed infrastructure asset class” that

was not already spanned by a combi-

nation of capital market instruments and

alternatives, or by a factor-based asset

allocation;

2. The majority of test portfolios that

improve the mean-variance efficient

frontier before the GFC fail to repeat this

feat post-GFC. There is no evidence of

persistent diversification benefits;

3. Of the 22 test portfolios used, only four

manage to improve on a typical asset

allocation defined either by traditional

asset class or by factor exposure after theGFC and only one is not already spanned

both pre- and post-GFC;

4. Building baskets of stocks on the basis of

their SIC code and sector-derived income

fails to generate a convincing exposure

to a new asset class.

5. Hence, benchmarking unlisted infras-

tructure investments with thematic

(industry-based) stock indices is unlikely

to be very helpful from a pure asset

allocation perspective i.e. the latter do

not exhibit a risk/return trade-off or

betas that large investors did not have

access to already.

Overall, we do not find persistent evidence

to support the claims that listed infras-

tructure is an asset class. In other words,

any ”listed infrastructure” effect was already

spanned by a combination of capital market

instruments over the past 15 years in Global,

US and UK markets.

Defining infrastructure investments as a

series of industrial sectors and/or tangible

assets is fundamentally misleading. We

find that such asset selection schemes do

not create diversification benefits, whether

reference portfolios are structured by tradi-

tional asset classes or factor exposures.

We conclude that what is typically referred

to as listed infrastructure, defined by SIC

code and industrial sector, is not an assetclass or a unique combination of marketfactors, but instead cannot be persistently

distinguished from existing exposures in



Executive Summary

Figure 1: Illustration of the difference tests of mean-variance spanning of the FTSEMacquarie USA Infrastructure Index

(a) 2000-2014, asset class and factor-based reference

●●

●

●

●

●

●

●

●

Government Bonds

Corporate Bonds

High Yield

Hedge Funds

Commodities

Real Estate

US Equities

World Ex−US Equities

FTSE Macquarie US Infra

−0.05

0.00

0.05

0.10

0.15

0.20

0.00 0.05 0.10 0.15 0.20 0.25 0.30Standard Deviation

Ret

urns

Reference Assets Reference and Infrastructure Assets

Efficient Frontier January 2000 to December 2013

●

●

●

●

●

●

Market

Size

Value

Term

Default


−0.05

0.00

0.05

0.10

0.15

0.20


Ret

urns


Efficient Frontier February 2007 to December 2013

(b) 2000-2008, asset class and factor-based reference

●

●

●

●

●

●

●

●

●

Government Bonds

Corporate Bonds

High Yield

Hedge Funds

Commodities

Real Estate

US Equities World Ex−US Equities


−0.30

−0.25

−0.20

−0.15

−0.10

−0.05

0.00

0.05

0.10

0.15

0.20

0.25

0.00 0.05 0.10 0.15 0.20 0.25 0.30 0.35 0.40Standard Deviation

Ret

urns



●

●

●

●

●

●

●

Europe Stocks

US Stocks

Size

Value

Term

Default

FTSE Macq

−0.05

0.00

0.05

0.10

0.15

0.20


Ret

urns



(c) 2009-2014, asset class and factor-based reference

●

●

●

●

●

●

●

●

●

Government Bonds

Corporate Bonds

High Yield

Hedge Funds

Commodities

Real Estate

US Equities



−0.05

0.00

0.05

0.10

0.15

0.20

0.25


Ret

urns



●

●

●

●

●

●

Europe Stocks

US Stocks

Size

Value

Term

FTSE Macq

−0.05

0.00

0.05

0.10

0.15

0.20


Ret

urns





Executive Summary

Figure 2: Mean-Variance Frontier of total returns PFI portfolio and reference portfolio, 2000-2014

●

●

●

●

●

●

●

UK Gilts

Real Estate

Hedge Funds

Commodities

UK Equities

World Ex−UK Equities

PFI Portfolio

−0.05

0.00

0.05

0.10

0.15

0.20


Ret

urns



investors’ portfolios, and that expecting the

emergence of a new or unique ”infras-

tructure asset class” by focusing on public

equities selected on the basis of industrial

sectors is unlikely to be very useful for

investors.

Figure 1 provides an illustration of these

results in the case of the FTSE Macquarie

Listed Infrastructure Index for the U.S.

market.

Thus, asset owners and managers who use

the common ”listed infrastructure” proxies

to benchmark private infrastructure invest-

ments are either misrepresenting (probably

over-estimating) the beta of private infras-

tructure, and usually have to include various

”add-ons” to such approaches, making them

completely ad hoc and unscientific.

Defining infrastructure differentlyOur tests also tentatively suggest a more

promising avenue to ”find infrastructure” in

the public equity space: focusing on under-

lying contractual or governance structures

that tend to maximise dividend payout and

pay dividends with great regularity, such

as the public-private partnerships (PPPs) or

master limited partnerships (MLPs) models,

we find that the mean-variance frontier of

a reference investor can be improved.

The answer to our initial question partly

depends on how ”infrastructure” is defined



Executive Summary

and understood as an asset selection

scheme.

Under our third definition of infrastructure,

which focuses on the relationship-specific

and contractual nature of the infrastructure

business, we find that listed infrastructure

may help identify exposures that have at

least the potential to persistently improve

portfolio diversification on a total return

basis, as figure 6 illustrates. This effect is

driven by the regularity and the size of

dividend payouts compared to other corpo-

rations, infrastructure or not.

What determines this ability to deliver

regular and high dividend payouts is the

contractual and governance structure of the

underlying businesses, not their belonging

to a given industrial sector. Bundles of PPP

project companies or MLPs behave differ-

ently than regular corporations i.e. their

ability to retain and control the free cash

flow of the firm is limited and they tend

to make large equity payouts. In the case if

PPP firms, as Blanc-Brude et al. (2016) show,

they also pay dividends with much greater

probability than other firms.

Going beyond sector exposures and

focusing on the underlying business model

of the firm is more likely to reveal a unique

combination of underlying risk factors.

However, it must be noted that the relatively

low aggregate market capitalisation of

listed entities offering a ”clean” exposure to

infrastructure ”business models” as opposed

to ”infrastructure corporates” may limit the

ability of investors to enjoy these potential

benefits unless the far larger unlisted infras-

tructure fund universe has similar charac-

teristics.

Future work by EDHECinfra aims to answer

these questions in the years to come.


1. Introduction



1. Introduction

In this paper, we ask the question: does

focusing on listed infrastructure stocks

create diversification benefits previously

unavailable to large investors already active

in public markets?

This question arises from what we call

the ”infrastructure investment narrative”

(Blanc-Brude, 2013), a set of investment

beliefs commonly held by investors about

the investment characteristics of infras-

tructure assets.

In this narrative, the ”infrastructure asset

class” is less exposed to the business cycle

because of the low price-elasticity of infras-

tructure services. Furthermore, the value of

these investment is expected to be mostly

determined by income streams extending

far into the future, and should thus be less

impacted by current events.

According to this intuition, infrastructure

investments may provide diversification

benefits to investors since they are expected

to exhibit low return covariance with other

financial assets, as well as a degree of

downside protection. In other words, infras-

tructure investments are expected to exhibit

sufficiently unique characteristics to be

considered an ”asset class” in its own right.

Empirically, there are at least three reasons

why this view requires further examination:

1. Most existing research on infrastructure

has used public equity markets to infer

findings for the whole infrastructure

investment universe, but robust and

conclusive evidence is not forthcoming

in existing papers;




invest in ”listed infrastructure” arguing

that it does indeed constitute an asset

class in its own right, worthy of an

individual allocation;

3. Listed infrastructure stocks are often


in privately-held (unlisted) infrastructure

but the adequacy of such proxies remains

untested.

The existence of a distinctive listed infras-tructure effect in investors’ portfolio would

support these views. In the negative, if

this effect cannot be found, there is little

to expect from listed infrastructure equity

from an asset allocation (risk/reward optimi-

sation) perspective and maybe even less

to learn from public markets about the

expected performance of unlisted infras-

tructure investments.

We test the impact of adding 22 different

proxies of public infrastructure stocks to the

portfolio of a well-diversified investor using

mean-variance spanning tests. We focus on

three definitions of ”listed infrastructure” as

an asset selection scheme:

1. A ”naïve”, rule-based filtering of stocks

based on industrial sector classifications

and percentage income generated from

pre-defined infrastructure sectors (nine

proxies);



1. Introduction

2. Existing listed infrastructure indices

designed and maintained by index

providers (twelve proxies);

3. A basket of stocks offering a pure

exposure to several hundred underlying

projects that correspond to a well-

known form of infrastructure investment

defined – in contrast with the two

previous cases – in terms of long-term

public-private contracts, not industrial

sectors (one proxy).

Overall, we do not find persistent evidence

to support the claims that listed infras-

tructure provides diversification benefits.

In other words, any ”listed infrastructure”

effect was already spanned by a combi-

nation of capital market instruments over

the past 15 years in Global, US and UK

markets.

We show that listed infrastructure, as it is

traditionally defined (by their SIC code and

industrial sector), is not an asset class or

a unique combination of market factors,

but instead cannot be distinguished from

existing exposures in investors’ portfolios.

We also find that the answer to our

question partly depends on how ”infras-

tructure” is defined and understood as an

asset selection scheme.

Hence, under our third definition of infras-

tructure, which focuses on the relationship-

specific and contractual nature of the

infrastructure business, we find that listed

infrastructure may help identify exposures

that have at least the potential to persis-

tently improve portfolio diversification. In

other words, going beyond sector exposures

and focusing on the underlying business

model of the firm is more likely to reveal

a unique combination of underlying risk

factors.

The rest of this paper is structured as

follows: section 2 briefly reviews existing

research on the performance of listed

infrastructure. Section 3 details our

approach, while sections 4 and 5 present

our choice of methodology and data,

respectively. The results of the analysis are

reported in section 6. Finally, Section 7

discusses our findings and their impli-

cations to better define and benchmark

infrastructure equity investments.


2. Literature Review




In Fabozzi and Markowitz (2011, p. 16),

asset classes are defined as homogenous

investments with comparable character-

istics, driven by similar factors, including a

common legal or regulatory structure, thus

correlating highly with each other.

As a direct result of this definition, the

combination of two or more asset classes

can be expected to create diversifi-

cation benefits due to the limited return

covariance of each group of assets.

The question of whether listed infras-

tructure is an asset class and a good proxy

of a broader universe of privately-held

infrastructure equity has been discussed in

previous research. The approach taken in the

literature has been to define infrastructure

in terms of industrial categories: roads and

airports can seem rather alike as businesses

when compared with automotive factories

or financial services. Hence, following the

definition given above, they can expected

to form a relatively homogeneous group of

stocks – a potential asset class – compared

to other segments of the economy.

Existing studies can be organised in two

groups. First, papers applying rule-based

stock selection schemes focusing on what is

traditionally understood as ”infrastructure”

i.e. a collection of industrial sectors. Second,

papers that employ listed infrastructure

indices created by a number of index

providers.

2.1 Rule-based listedinfrastructure portfoliosThe first group of papers simply examines

those stocks that are classified under a set

list of infrastructure activities and derive

a certain proportion of their income from

these activities (see Newell and Peng (2007),

Finkenzeller et al. (2010), Newell and Peng

(2008), Newell et al. (2009), Rothballer and

Kaserer (2012a) and Bitsch (2012)).

The findings from these studies suggest

considerable heterogeneity in ”listed infras-

tructure”. Newell and Peng (2007) report

that listed Australian infrastructure exhibits

higher returns, but also higher volatilitythan equity markets. They find higher

Sharpe ratios than the market and low but

growing correlations over time with market

returns. Finkenzeller et al. (2010) find similar

results.

The work of Newell and Peng (2008)

finds that in the U.S., infrastructure(ex-utilities) under-performs stocks and

bonds over the period from 2000 to

2006, while utilities outperform the market.

Rothballer and Kaserer (2012a) find that

infrastructure stocks exhibit lower marketrisk than equities in general but notlower total risk i.e. they find high idiosyn-

cratic volatility.

They also report significant heterogeneity in

the risk profiles of different infrastructure

sectors with an average beta of 0.68 but

with variation between sectors. For the

utility, transport and telecom companies,

the average betas were 0.50, 0.73 and 1.09,




respectively. 1 Bitsch (2012) finds that infras-1 - Using the same sample thanRothballer and Kaserer (2012a), Rödeland Rothballer (2012), examine theinflation hedging ability of infras-tructure.They find no evidence to suggestinfrastructure exhibits a greaterability to hedge inflation risks thanlisted equities. Even restricting theirsample to firms with assumed strongmonopoly characteristics does notyield a statistically significant result.

tructure vehicles are priced using a high risk

premium in part because of – he argue –

complex and opaque financial structuring,

information asymmetries with managers,

regulatory and political risks.

These findings are in line with the results

of several industry studies suggesting that

the volatility of infrastructure indices is on

par with equities and real estate, but that

market correlation is relatively low (Colonial

First State Asset Management, 2009; RREEF

2007).

The conclusions from this strand of liter-

ature are limited. ”Infrastructure” stocks are

founds to have higher Sharpe ratios in

some cases but the statistical significanceof this effect is never tested. Overall,

rule-based infrastructure stocks selection

schemes lead to either anecdotal (small

sample) or heterogenous results, which do

not support the notion of an independent

asset class.

2.2 Ad hoc listed infrastructureindicesA second group of studies uses infras-

tructure indices created by index providers

such as Dow Jones, FTSE, MSCI and S&P,

as well as financial institutions such as

Brookfield, Macquarie or UBS. These indices

are not fundamentally different from the

approach described above.

They use asset selection schemes based on

slightly different industrial definitions of

what qualifies as ”infrastructure” and apply

a market-capitalisation weighing scheme.

They are ad hoc as opposed to rule-based

because index providers pick and choose

which stocks should be included in each

infrastructure index.

Using such indices, Bird et al. (2014) and

Bianchi et al. (2014) find that infrastructure

exhibits similar returns, correlations and

tail-risks than the market, with a marginally

higher Sharpe ratio, driven by what could be

described as a ’utility tilt’.

Other recent studies on the performance of

infrastructure indices by Peng and Newell

(2007), Finkenzeller et al. (2010), Dechant

and Finkenzeller (2013) and Oyedele et al.

(2014) also report potential diversification

benefits, but none examine whether these

are statistically or economically significant.

For example, Peng and Newell (2007) and

Oyedele et al. (2014) compare Sharpe ratios

but provide no statistical tests to supporttheir conclusions.

Idzorek and Armstrong (2009) provide the

only study of the role of listed infrastructure

in a portfolio context. The authors create an

infrastructure index by combining existing

industry indices. Using three versions of

their composite index (low, medium and

high utilities), and consistent with previous

papers, they report that over the 1990-

2007 period, infrastructure returns were

similar to that of U.S. equities but with

slightly less risk. Finally, using the CAPM

to create a forward-looking model of

expected returns including an infrastructure




allocation, Idzorek and Armstrong (2009)

find that adding infrastructure does not

lead to a meaningful improvement of the

efficient frontier, but again provide no

statistical test of the robustness of their

results.

2.3 Limitations of existing researchThe existing literature has not examined

whether different types of listed infras-

tructure investments are not already

spanned in the portfolio of a typical

investor. As a result, it remains unclear

whether a focus on infrastructure-related

stocks can create additional diversification

benefits for investors. Nor is it clear whether

”infrastructure” is a new combination of

investment factor exposures.

In the rest of this paper, we test statisti-

cally whether infrastructure stocks, selected

according to their industrial classification,

provide diversification benefits to investors.

Furthermore, following the argument in

Blanc-Brude (2013), we examine a different

definition of infrastructure focussing on

the ”business model” as determined by the

role of long-term contracts in infrastructure

projects. Next, we describe our approach in

more detail.


3. Approach



3. Approach

3.1 Testing three definitions oflisted infrastructureWe propose to test the portfolio character-

istics of listed infrastructure equity under

the three different definitions of what

constitutes ”infrastructure” proposed in

section 1.

These first two proxies – a ”naïve”, rule-

based filtering of stocks based on industrial

sector classifications and listed infras-

tructure indices maintained by index

providers – focus on the ”real” charac-

teristics of the relevant capital projects

and bundles together assets that may

all be related to large structures of steel

and concrete, but may also have radically

different risk profiles from an investment

view point.

Hence, we also identify stocks which happen

to create a useful natural experiment 2: a2 - i.e. Experimental and controlconditions are determined by factorsoutside our control

basket of stocks offering a pure exposure to

projects that correspond to a specific long-

term contract but not to any specific indus-

trial sectors.

Instead, this basket captures a specific

infrastructure ”business model”: these are

the publicly traded shares of investment

vehicles that are solely involved in buying

and holding the equity of infrastructure

projects engaged in PFI (Private Finance

Initiative) projects in the UK and, to a lesser

extent, their equivalents in Canada, France

and the rest of the OECD.

PFI projects consists of dedicated project

firms entering into long-term contracts

with the public sector to build, maintain

and operate public infrastructure facilities

according to a pre-agreed service output

specification. As long as these firms deliver

the projects and associated services for

which they have been mandated, on time

and according to specifications, the public

sector is committed to pay a monthly or

quarterly income to the firm according to

a pre-agreed schedule for multiple decades.

In the UK, the long-term contract between

the public and private parties also stipu-

lated that this ”availability payment” is also

adjusted to reflect changes in the retail

price index (RPI). Each project company

is a special purpose vehicle created solely

to deliver an infrastructure project and

financed on a non-recourse basis with

sponsor equity, shareholder loans and senior

debt.

The firms we identify are listed on the

London Stock Exchange, buy and hold the

equity and shareholder loans (quasi-equity)

of hundreds of these PFI project companies

and subsequently distribute dividends to

they shareholders. They are, in effect, a listed

basket of PFI equity (with no additional

leverage) and, as such represent a pure

proxy of one model of private infrastructure

investment. We provide a full list of the

underlying firms in the appendix.

We expect the cash flows of these firms

to be highly predictable, uncorrelated with

markets and the business cycle, albeit highly

correlated with the UK RPI index. In other

words, we expect to see some evidence

of the ”infrastructure investment narrative”



3. Approach

discussed earlier, that has so far eluded

studies of infrastructure stocks defined by

their SIC code.

3.2 Testing asset classes orfactors?Using these three alternative approaches to

define infrastructure investment, we use the

mean-variance spanning tests designed by

Huberman and Kandel (1987) and Kan and

Zhou (2012) to determine whether adding

a listed infrastructure bucket to an existing

investment portfolio significantly increases

diversification opportunities.

In the affirmative, this result implies a

degree of ”asset class-ness” of infras-

tructure stocks since their addition to a

reference portfolio effectively shifts the

mean-variance efficient frontier (to the left)

and creates new diversification opportu-

nities for investors. Furthermore, we define

the reference portfolio used to test the

mean-variance spanning properties of listed

infrastructure either in terms of traditional

asset classes and of investment factors.

Indeed, the notion of asset class is losing its

relevance in investment management since

the financial crisis of 2008-11, when existing

asset class-based allocations failed to prove

well diversified (see for example Ilmanen

and Kizer, 2012).

Factor-based asset allocations aim to

identify those persistent dimensions of

financial assets that best explain (and

predict) their performance instead of

assuming that assets belong to distinctive

categories because they have different

names. 33 - Such factors include the Famaand French (1992) Size and Valuepremiums, the Term and Defaultpremiums (Fama and French, 1993)and the Momentum premiumidentified by Jegadeesh and Titman(1993). Bender et al. (2010) showsthat these premiums are uncorrelatedwith each other, increase returnsand reduce portfolio volatility overtraditional asset class allocations.Likewise, when comparing the diver-sification benefits of factor-basedallocations to alternative assets,Bird et al. (2013) finds that factorapproaches tend to outperformalternative asset classes. For recentand in-depth analyses of factorinvesting, see Amenc et al. (2014).

Thus, we include both a traditional asset

allocation based on asset classes and a

factor based allocation to test whether

listed infrastructure is indeed an asset class

or, alternatively, a unique combination of

investment factors.

3.3 Testing persistenceFinally, we test the existence of a persistenteffect of listed infrastructure on a reference

portfolio by splitting the observation period

in two, from 2000 to 2008 and from 2008

to 2014, to test for the impact of the 2008

reversal of the credit cycle a.k.a. the Global

Financial Crisis (GFC) and the distorting role

excessive leverage may have played in the

first period. In section 6, we report results

for the whole sample period, as well as for

two sub-sample periods denoted as pre- and

post-GFC periods.

In the case of the PFI portfolio, data starts

in 2006 and so we also split the sample in

2011, which marks the time of the Eurozone

debt crisis and launch of quantitative easing

policies by the Bank of England.

In the next section, we discuss the mean-

variance spanning methodology used in the

remaining sections of this paper.


4. Methodology



4. Methodology

In a mean-variance framework, the

question of whether listed infrastructure

provides diversification benefits is equiv-

alent to asking whether investors are able

to improve their optimal mean-variance

frontier by including infrastructure stocks

to an existing portfolio.

This question can be answered using the

mean-variance spanning test described by

Huberman and Kandel (1987), which tests

whether the efficient frontier is improved

up to given level of statistical signifi-cance when including new assets.

If the mean-variance frontier, inclusive of

the new assets, coincides with that already

produced by the reference assets, the new

assets can be considered to be already

spanned by the existing portfolio i.e. no new

diversification benefit is created. Conversely,

if the existing mean-variance frontier is

shifted to the left in the mean/variance

plane, by the addition of the new asset,

investors have improved their investment

opportunity set.

This approach has been used to examine

the diversification benefits in different asset

classes. For instance, Petrella (2005) and Eun

et al. (2008) employ this methodology to

examine the diversification benefits of small

cap stocks. Likewise, Kroencke and Schindler

(2012) examines the benefits of interna-

tional diversification in real estate using

mean-variance spanning, while Chen et al.

(2011) examines the diversification benefits

of volatility. However, to date it has not

been used in the literature on listed infras-

tructure.

Mean-variance spanning is a regression

based test that assumes that there are Kreference assets as well as N test assets.

In Huberman and Kandel (1987), there

is a linear relationship between test and

reference assets so that:

R2t = α + βR1t + εt (4.1)

with t = 1, . . . T periods and R1 represents

a T × K matrix of excess realised returns

of the K benchmark assets. R2 represents a

T × N matrix of realised excess returns of

the N test assets. β is a K × N matrix of

regression factor loadings and ε is a vector

of the regression error terms.

The null hypothesis is that existing assets

already span the new assets. This implies

that the α of the regression in equation 4.1

is equal to zero, whilst the sum of the βs

equals one. As a result, the null hypothesis

assumes that a combination of the existing

benchmark assets is capable of replicating

the returns of the test assets with a lower

variance.

Kan and Zhou (2012) describes the null

hypothesis as :

H0S = α = 0N, δ = 1N − β1K = 0N(4.2)

Where αN is an N-vector of regression

intercept coefficients and β is the matrix of

factor loadings.



4. Methodology

As this analysis is only examining the

addition of one test asset at a time the test

statistic is given by: 4

4 - Kan and Zhou (2012) state thatif N ≥ 2 then the appropriateformation of the test statistic is given

as HK =

(1

V12

− 1)(

T−K−12

).

HK =(

1V− 1

)(T− K− 1

2

)(4.3)

where V is the ratio of the determinant

of the maximum likelihood estimator of

the error co-variance matrix of the model

assuming that there is no spanning of

the efficient frontier (otherwise known as

the unrestricted model) to that of the

determinant of the maximum likelihood

estimator of the model that assumes

spanning occurs (known as the restricted

model).

T is the number of return observations, K is

the number of benchmark assets included

in the study. The HK variable is a Wald-test

statistic and follows an F-distribution with

(2, T− K− 1) degrees of freedom.

In addition to the Huberman and Kandel

(1987) test, Kan and Zhou (2012) develop

a two-stage test to examine whether the

rejection of the Huberman and Kandel

(1987) null hypothesis is due to differences

in the tangency or the Global Minimum

Variance as a result of the addition of new

assets.

The first step of the Kan and Zhou (2012)

test examines whether α = 0N. If the null

is rejected at this stage, the two tangency

portfolios comprising of the benchmark

assets, and the benchmark and new assets,

respectively, are statistically different.

The second stage of the Kan and Zhou

(2012) test examines whether δ = 0Nconditional on α = 0N. If the null

hypothesis is rejected, the Global Minimum

Variance of the test portfolio and the

benchmark portfolios are statistically

different (see also Chen et al., 2011, for a

discussion).

In this paper, we incorporate both the

Huberman and Kandel (1987) and Kan and

Zhou (2012) tests to examine the ability of

infrastructure to provide portfolio diversifi-

cation benefits. 55 - As another robustness check weemploy the Gibbons et al. (1989) testof portfolio efficiency. The results, aresimilar to the mean-variance spantest results presented in this paper.The results are available upon request.

Next, we describe the data used in this study.


5. Data



5. Data

This section describes the datasets used to

build test portfolios of listed infrastructure

and reference portfolios to apply the mean-

variance spanning methodology described

previously.

Sections 5.1, 5.2, and 5.3 describe listed

infrastructure proxies designed with sector-

based asset selection rules, index-provider

data, and the PFI portfolio, respec-

tively. Section 5.4 describes the reference

portfolios.

5.1 Test Assets - Listedinfrastructure companies5.1.1 Asset selectionThe first asset selection scheme repre-

sents the ”naïve” definition of infrastructure

equity investment, and follows the method-

ology described by Rothballer and Kaserer

(2012b) following broad industry definitions

to determine infrastructure-related stocks 6.6 - The SIC and GIC codes used toidentify infrastructure are availableupon request.

5,757 possible securities listed in

global markets are thus identified as

infrastructure-related. Next, only stocks

for which the majority of the revenue was

obtained from sectors corresponding to

infrastructure activities are kept in the

sample.

A minimum market capitalisation of

USD500 Million is also required to be

included in the sample. This yields 1,290

firms with at least 50% of their income

from infrastructure activities The minimum

revenue by infrastructure type is reported

by SIC or GIC code by Worldscope. This

is a crude measure as it relies on the

continuous updating of the revenue codes

by Worldscope, as well as assuming that

GIC or SIC codes represent infrastructure

activities.

Setting a minimum infrastructure sector

revenue threshold to 75% and 90%, yields

650 and 554 stocks, respectively.

U.S. dollar price and total returns are

sourced from Datastream using the

methodology described in Ince and Porter

(2006). 7 The firms thus identified comprise

7 - Extreme monthly returns areidentified following Ince and Porter(2006) and set to a missing value.Ince and Porter (2006) sets an arbitarycut off of 300% for extreme monthlyreturns. If R1 or Rt−1 is greater than300% and (1+R1)/(1+Rt−1)−1is less than 50% then R1 or Rt−1 areset to missing. Furthermore, followingRothballer and Kaserer (2012b), 18months of non zero returns arerequired for the stock to be includedin the portfolios.Any Datastream padded price isremoved by requesting X(P#S) $Uwhich returns null values whenDatastream does not have a recordand any non equity item is removedby requiring the TYPE description inDatastream to be equal to EQ.

at most 12%, 7% and 6.5% of the of

the MSCI World market value as at 31

December 2014, for the 50%, 75% and

90% infrastructure revenue thresholds,

respectively.

5.1.2 Descriptive StatisticsFor market-cap weighted portfolios of

infrastructure stocks defined according

to the industry-based scheme described

above, we report annualised returns,

standard deviation and Sharpe ratios and

the maximum drawdown ratios for the

period 2000-2014 for price and total

returns. in table 1.

It is useful to note that the reference

market index should not be difficult to

beat. The MSCI World index is a free float

adjusted market capitalisation weighted

index comprising of 1,631 mid and large

capitalisation stocks across 23 developed

country equity markets. MSCI states

that the index comprises 85% of the

free-float adjusted market capitalisation



5. Data

Table 1: This table presents the descriptive statistics for the naïve infrastructure stockselection scheme, 2000-2014. Telecom, Transport and Utilities are portfolios of stocks thatearn a minimum revenue level from activities related to SIC or GIC codes recognised asTelecommunications, Transport and Utilities industries, respectfully. The minimum revenuelevels required are 50%, 75% and 90% , respectfully. Return is the average monthly returnfrom January 2000 to December 2014. Risk is the monthly standard deviation of returnsfrom January 2000 to December 2014. SR is the Sharpe Ratio calculated with the averageyield of the U.S. one month Treasury bill as the risk free rate proxy. Worst Drawdown is themaximum drawdown ratio, measured as a percentage of maximum cumulative return i.e.from ”peak equity.

Price returns, 50% Rev Tel. 50% Rev Transp. 50% Rev Util. MSCI WorldPrice return −0.084 0.020 −0.005 0.012Risk 0.179 0.154 0.152 0.158SR −0.496 0.094 −0.065 0.047Worst Drawdown 0.830 0.620 0.570 0.550, 75% Rev Tel. 75% Rev Transp. 75% Rev Util. MSCI WorldPrice return −0.092 0.068 0.003 0.012Risk 0.185 0.198 0.133 0.158SR −0.522 0.318 −0.015 0.047Worst Drawdown 0.830 0.690 0.500 0.550, 90% Rev Tel. 90% Rev Transp. 90% Rev Util. MSCI WorldPrice return −0.085 0.042 0.002 0.012Risk 0.175 0.180 0.133 0.158SR −0.515 0.205 −0.025 0.047Worst Drawdown 0.810 0.690 0.480 0.550

Total returns, 50% Rev Tel. 50% Rev Transp. 50% Rev Util. MSCI WorldTot. return −0.052 0.048 0.028 0.036Risk 0.180 0.152 0.153 0.159SR −0.315 0.282 0.148 0.197Worst Drawdown 0.818 0.609 0.556 0.537, 75% Rev Tel. 75% Rev Transp. 75% Rev Util. MSCI WorldTot. return −0.057 0.109 0.039 0.036Risk 0.185 0.197 0.134 0.159SR −0.334 0.523 0.255 0.197Worst Drawdown 0.826 0.667 0.473 0.537, 90% Rev Tel. 90% Rev Transp. 90% Rev Util. MSCI WorldTot. return −0.051 0.088 0.038 0.036Risk 0.176 0.179 0.133 0.159SR −0.315 0.461 0.249 0.197Worst Drawdown 0.799 0.662 0.454 0.537



5. Data

of each country covered. The index is

updated quarterly with annual revisions

to update the investable universe and the

removal of stocks with low liquidity. Such

market value-weighted indices, while they

constitute a useful point of reference, have

been shown to be highly inefficient in

previous research (see Amenc et al., 2010).

Nevertheless, the listed infrastructure

portfolios obtained above do not necessarily

offer better risk-adjusted performance than

this relatively unambitious baseline.

We observe that irrespective of the revenue

cut-offs employed to form the infras-

tructure portfolios, the telecom sector

continually produces poor returns. This

sector does not seem to have recovered

from the technology bubble of the yearly

2000s. This suggests that certain ”infras-

tructure” sectors experience a high degree

of cyclicality, as well as a complete absence

of persistence. Transportation fares better

with higher Sharpe ratios than the market

index under both the price return and

total return measures. However, drawdown

risk is typically higher than the market’s

for price and total returns as shown in

Table 1. During the sample period Utilities

only outperform the broad market from a

total return perspective.

Next, we describe our second test asset,

a combination of rule-based and ad hocstock selection schemes created by index

providers.

5.2 Test assets - Ad hoc listedinfrastructure indices5.2.1 Asset selectionThe basic requirements to be included in

listed infrastructure indices created by index

providers are not very different from the

naïve selection scheme described above.

They include:

1. being part of a broader index universe

(usually that of the infrastructure

universe of the index provider); and,

2. a minimum amount of revenue derived

from infrastructure activities.

However, minimum revenue requirements

and the definition of infrastructure activ-

ities are set differently by each index

provider, adding what could amount to

”active views”, to a rule-based scheme.

We test two groups of listed infrastructure

indices: a set of global indices and one

designed to represent the U.S. market only.

Global indices provide a direct comparison

with the naïve approach described above,

while a U.S.-only perspective allows more

controls and granularity when designing a

reference portfolio of asset classes or factors

to test the mean-variance spanning of listed

infrastructure indices.

Global Infrastructure IndicesWe include seven global infrastructure

indices and four U.S. infrastructure indices: 88 - A brief summary of the indices isavailable upon request

l Dow Jones Brookfield Global Infras-

tructure Index;



5. Data

l FTSE Macquarie Global Infrastructure

Index;

l FTSE Global Core Infrastructure;

l MSCI World Infrastructure Index;

l MSCI ACWI Infrastructure Capped;

l UBS Global Infrastructure and Utilities;

and,

l UBS Global 50/50 Infrastructure and

Utilities.

The universe thus recognised by index

providers is not very large with only the

MSCI World Infrastructure and MSCI ACWI

Global Infrastructure representing more

than 10% of the value of the MSCI World

Index.

U.S. infrastructure indicesThe U.S. infrastructure indices included in

this study are:

l FTSE Macquarie USA Infrastructure Index;

l MSCI US Infrastructure Index;

l MSCI USA Infrastructure 20/35 Capped

Index; and,

l Alerian MLP Infrastructure Index.

5.2.2 Descriptive statisticsGlobal Infrastructure Indices

Table 2 shows that most infrastructure

indices have higher Sharpe ratios than

the reference market index (MSCI World).

The Dow Jones Brookfield Global Infras-

tructure index exhibits the highest average

annualised returns and Sharpe ratio for

the sample period. This performance is

contrasted by theMSCIWorld Infrastructure

index which exhibits negative performance

on a price return basis. However Table 2

suggest that drawdown risk is very similar

between infrastructure indices and the

broad market, with the exception of the

Brookfield and MSCI ACWI indices.

U.S. infrastructure indices

In the case of US-only indices, the MSCI

and FTSE indices, reported in Table 3 do

not seem to stand out from the broad

market index (here the Russell 3000), but the

Alerian MLP index, which captures a under-

lying business model focused on dividend

distributions, exhibits very different charac-

teristics. with much higher Sharpe ratios

especially on a total return basis, but equally

high maximum drawdown.

5.3 Test Assets - Listed baskets ofcontracted infrastructure projects5.3.1 Asset selectionThe PFI portfolio consists of

1. HSBC Infrastructure Company Ltd (HICL)

2. John Laing Infrastructure Fund Ltd (JLIF)

3. GCP Infrastructure Ltd (GCP)

4. International Partnerships Ltd (INPP)

5. Bilfinger Berger Global Infrastructure Ltd

(BBGI)

As discussed, these firms are solely

occupied with buying and holding the

equity and quasi-equity of PFI (private

finance initiative) project companies in

existence in the U.K. and that of similar



5. Data

Table 2: This table presents the descriptive statistics for the Global Infrastructure Indicesfor the period, 2000-2014. BF is the Dow Jones Brookfield Global Infrastructure Index, SP isStandard & Poor’s Global Infrastructure Index, FTSEM is the FTSE Macquarie Global Infras-tructure Index, FTSEC is the FTSE Global Core Infrastructure Index, MSCI is the MSCI WorldInfrastructure Index, MSCIA is the MSCI ACWI Infrastructure Capped, UBS is the UBS GlobalInfrastructure and Utilities, UBS 50 is the UBS Global 50/50 Infrastructure and Utilities Indexand MSCIW is the MSCI World Index. Return is the average monthly return from the indexcommencement date to December 2014. Risk is the monthly standard deviation of returnsfrom the index commencement date to December 2014. SR is the Sharpe Ratio calcu-lated with the average yield of the U.S. one month Treasury bill as the risk free rate proxy.Worst Drawdown is the maximum drawdown ratio, measured as a percentage of maximumcumulative return i.e. from ”peak equity”.

Price returns, BF SP FTSEM FTSEC MSCI MSCIA UBS UBS50 MSCIWReturn 0.112 0.072 0.043 0.063 −0.020 0.025 0.046 0.055 0.012Risk 0.132 0.153 0.138 0.115 0.145 0.105 0.119 0.145 0.158SR 0.807 0.436 0.273 0.507 −0.171 0.192 0.347 0.341 0.047Worst Drawdown 0.476 0.551 0.456 0.374 0.660 0.424 0.447 0.505 0.554

Total returns, BF SP FTSEM FTSEC MSCI MSCIA UBS UBS50 MSCIWReturn 0.147 0.116 0.083 0.099 0.019 0.061 0.082 0.091 0.036Risk 0.132 0.153 0.138 0.114 0.146 0.105 0.119 0.145 0.159SR 1.070 0.725 0.561 0.820 0.092 0.532 0.644 0.588 0.197Worst Drawdown 0.452 0.527 0.432 0.348 0.640 0.395 0.426 0.484 0.537

Table 3: This table presents the descriptive statistics for of annualised price and total returnsof U.S. infrastructure stock indices, 2000-2014. AMLP is the Alerian MLP Infrastructure Index,FTSEM is the FTSE Macquarie USA Infrastructure Index, MSCI is the MSCI US InfrastructureIndex, MSCISC is the MSCI USA Infrastructure 20/35 Capped Index and R3000 is the Russell3000 index. Return is the average monthly return from January 2000 to December 2014.Risk is the monthly standard deviation of returns from January 2000 to December 2014. SRis the Sharpe Ratio calculated with the average yield of the U.S. one month Treasury bill asthe risk free rate proxy. Worst Drawdown is the maximum drawdown ratio, measured as apercentage of maximum cumulative return i.e. from ”peak equity”.

Price returns, AMLP FTSEM MSCI MSCISC R3000Price return 0.130 0.063 −0.016 0.024 0.029Risk 0.166 0.448 0.147 0.138 0.157SR 0.748 0.130 −0.141 0.133 0.155Worst Drawdown 0.492 0.956 0.633 0.448 0.527

Total returns, AMLP FTSEM MSCI MSCISC R3000ann. total return 0.213 0.067 0.021 0.055 0.048ann. risk 0.170 0.448 0.148 0.138 0.157ann. Sharpe ratio 1.219 0.137 0.106 0.361 0.274Worst Drawdown 0.431 0.956 0.609 0.423 0.512



5. Data

firms mostly involved in delivering so-

called availability-payment infrastructure

projects, by which the public sector pays

a pre-agreed income to the project firm

on a regular basis in exchange for the

construction/development, maintenance

and operations of a given infrastructure

project given a pre-agreed output specifi-

cation and for several decades.

These PFI project companies in question

do not enter into any other activities

during their lifetime, and solely deliver the

contracted infrastructure and associated

services while repaying their creditors and

investors. As such, they give access to

a “pure” infrastructure project cash flow,

representative of the underlying nature of

the PFI business model.

The firms in the PFI portfolio can be

considered useful proxies of a portfolio of

PFI equity investments. While the project

companies are typically highly leveraged,

the firms in the PFI portfolio do not make

a significant use of leverage. Hence, as a

group, they can be considered to be repre-

sentative a listed basket of PFI equity stakes.

All returns are annualised monthly price and

total returns computed in local currency

(GBP) and sourced from Datastream.

5.3.2 Descriptive StatisticsTable 4 suggest that the PFI portfolio

possesses different characteristics than

other listed infrastructure portfolios

examined so far. Its Sharpe ratio is high

but its maximum drawdown is much lower

than the market reference (here the FTSE All

Shares). Indeed, the maximum drawdown

for the PFI portfolio is also much lower

than the FTSE Macquarie Europe infras-

tructure index, another listed infrastructure

index focused solely on European markets.

The combination of high risk-adjusted

performance with low drawdown risk is

particularly striking in the total return case.

5.4 Reference AssetsAs discussed above, we use two types of

reference allocations to test the impact of

adding listed infrastructure to an investor’s

universe, an asset class-based allocation and

a factor-based allocation. All the summary

statistics for the reference assets can be

found in table 13 in the Appendix.

5.4.1 Global asset class-basedreference portfolioA ”well diversified investor” in the traditional

albeit imprecise meaning of the term can be

expected to hold a number of different asset

classes, including:

l Global Fixed Interest proxied by JP

Morgan Global Aggregate Bond Index;

l Commodities proxied by The S&P

Goldman Sachs Commodity Index;

l Real Estate proxied by MSCI World Real

Estate Index;

l Hedge Funds proxied by Dow Jones Credit

Suisse Hedge Fund Index; and,

l OECD and Emerging Market Equities

proxied by MSCI World and MSCI

Emerging Market Indices, respectively.



5. Data

Table 4: This table presents the descriptive statistics for of annualised price and total returnsof the PFI portfolio, an infrastructure index and the market index, 2006-2014. The PFIPortfolio is the equal-weighted return of the PFI stocks identified listed on the London StockExchange. Return is the average monthly return from 2006 to December 2014. Risk is themonthly standard deviation of returns from 2006 to December 2014. SR is the Sharpe Ratiocalculated with the average yield of the U.S. one month Treasury bill as the risk free rateproxy. Max. DD is the maximum drawdown ratio, measured as a percentage of maximumcumulative return i.e. from ”peak equity”. All returns are annualised monthly price and totalreturns computed in local currency (GBP) and sourced from Datastream.

Price returns, PFI Portfolio FTSE All Shares Macquarie Infra EuropePrice return 0.048 0.027 −0.007Risk 0.093 0.182 0.181SR 0.460 0.121 −0.065Max. DD 0.240 0.450 0.500

Total returns, PFI Portfolio FTSE All Shares Macquarie Infra EuropeTot. return 0.101 0.065 0.046Risk 0.082 0.172 0.184SR 1.171 0.345 0.224Max. DD 0.150 0.410 0.370

One potential issue with employing indices

as a reference asset is the possibility of

double-counting infrastructure stocks in

both the reference and test assets. This has

the potential of biasing the mean-variance

span tests against finding an improvement

in the investment opportunity set. Whilst

ideally removing any infrastructure like

stocks from the reference assets would solve

the problem of double counting, the circu-

lation of index constituent lists is too limited

to allow this.

However, the MSCI World Index, MSCI

(2014) states that as at November 2014 the

Utilities and Telecom industries comprise

3.32% and 3.46% whilst the Industrials

sector comprises 10.89% and the share of

”infrastructure” in industrials (e.g. railway) is

small.

Whilst it would be preferable to exclude the

infrastructure stocks from the MSCI World,

we cannot know the constituents and so

this cannot be done.

Nevertheless, given the low weighting

its likely that any results will not be

biased against the infrastructure stocks. We

conclude that not isolating infrastructure

stocks from our reference assets will not

materially influence the conclusions of this

study.

5.4.2 U.S. asset class referenceportfolioA typical U.S.-based reference portfolio

built using traditional asset classes would

include:

l Government Bonds proxied by the

Barclays Govt Aggregate Index;

l Corporate Bonds represented by the

Barclays U.S. Aggregate Index;

l High Yield Bonds with the Barclays U.S.

Corporate High Yield;



5. Data

l Real Estate, as per the US-DataStream

Real Estate Index;

l Hedge Funds represented by the Dow

Jones Credit Suisse Hedge Fund Index;

l Commodities proxied by the S&P


l U.S. Equities captured by the RUSSELL

3000 9; and,9 - The Russell 3000 index wasselected for the U.S. equity marketindex for two reasons. Firstly, it repre-sents the top 3,000 stocks by marketcapitalisation. This represents a signif-icant proportion of the investableuniverse of U.S. stocks. Secondly, forconsistency, in the factor exposurestudies we employ the Russell indicesto create the factor proxies.

l World Equities represented by the MSCI

World ex-US.

5.4.3 U.K. asset class referenceportfolioTo test the mean variance spanning

properties of the PFI portfolio, we build

a U.K. asset class reference portfolio

consisting of:

l Fixed Interest, represented by the Bank of

America/ML U.K. Gilts index;

l Real Estate, proxied by the DataStream

U.K. Real estate Index;

l Hedge funds, represented by the U.K.

DataStream hedge funds Index;

l Commodities, as proxied by the S&P


l U.K. Equities represented by the FTSE100;

and.

l World Equities proxied by the FTSE World

ex-U.K.

5.4.4 Global factor-based referenceportfolioConsistent with prior research, the factors

in this study are constructed from stock and

bond market indices. We follow Bender et al.

(2010), Ilmanen and Kizer (2012) and Bird

et al. (2013) to build Market, Size, Value,

Term and Default factors.

l The Market factor is the excess return of

the MSCI U.S. and MSCI Europe indices.

l The Size factor (SMB) is calculated by

taking the difference between the simple

average of MSCI Small Value and Growth

indices and the simple average of MSCI

Large Value and Growth Indices.

l The Value factor (HML) is constructed by

obtaining the difference between simple

average of MSCI Small, Mid and Large

Value indices and simple average of MSCI

Small, Mid and Large Growth Indices.

l The Term factor is estimated by taking the

difference between the returns of the U.S.

Government 10 year index and S&P U.S.

Treasury Bill 0-3 Index.

l Finally, the Default factor is estimated by

the change in the Moody’s Seasoned Baa

Corporate Bond Yield Relative to the Yield

on 10-Year Treasury Constant Maturity.

5.4.5 U.S. factor-based referenceportfolioU.S. factors are computed using the nowcanonical formulas reported in (Faff, 2001):

Market = Russell 3000 Index

−US one month Treasury Bill return

SMB =(Russell2000Value+ Russell2000Growth)

2

− (Russell1000Value+ Russell1000Growth)2

HML =(Russell1000Value+ Russell2000Value)

2

− (Russell1000Growth+ Russell2000Growth)2

Term = Barclays US Treasury 10-20 years Index

−Barclays US Treasury Bills 1-3 months Index

Default = Barclays US Corporate: AAA Long Index

−Barclays US Treasury Long Index



5. Data

In the next section, we present the results

of the mean-variance spanning tests

presented in section 4 using the multiple

datasets described above.


6. Results



6. Results

We first present the results of the Mean-

Variance Spanning tests conducted using

the asset classes as the reference assets in

section 6.1. Next, we use the factors defined

above as the reference portfolio.

We report the Huberman and Kandel (1987)

regression results for equation 4.1 and the

corresponding Kan and Zhou (2012) two

stage, step down tests.

6.1 Asset class mean-variancespanning test results6.1.1 Listed infrastructure companiesThe results of the mean-variance span test

for the naïve infrastructure portfolios are

reported in Table 5. For the price returns

of the nine portfolios constructed, Table 5

shows that the reference investment

opportunity set is improved by four of these

portfolios. These are the 75% Revenue

Cutoff Transport portfolio as well as the

Telecommunication portfolios. However,

when total returns are examined, only

the 50% Telecoms and the 75% Transport

infrastructure portfolio are found to

reject the Huberman and Kandel (1987)

null hypothesis of spanning at conven-

tional significance levels. Every otherportfolio does not improve upon themean-variance frontier created by thereference asset classes.

Applying the Kan and Zhou (2012) method-

ology, the results in Table 5 shows thatnone of the infrastructure portfoliosimprove the mean-variance frontierfrom that created by the reference

investments. Indeed, the listed infras-

tructure portfolios fail to improve either

the tangent portfolio or produce a lower

minimum variance portfolio. This finding is

consistent when either price or total returns

are used.

When the sub-periods are considered, both

before and after the GFC, the conclusion

that the naïve infrastructure approach fails

to identify any diversification benefits is

supported.

Panels B and D in Table 5 present the results

for the Mean-Variance Spanning tests for

the period January 2000 to December 2008.

The Huberman and Kandel (1987) test’s null

hypothesis is rejected in two cases: the

price and total returns of the 75% Transport

portfolio. However, the Kan and Zhou (2012)

test fails to reject the null hypothesis thatthe reference portfolio already spanslisted infrastructure.

From January 2009 to December 2014 (Panel

C and F in Table 5) only one portfolio is found

to improve the efficient frontier, the price

returns of the 50% Utilities portfolio. Here,

the Huberman and Kandel (1987) portfolio’s

null hypothesis is rejected at the 5% signif-

icance level and both steps of the Kan and

Zhou (2012) test reject the null hypothesis

that the portfolio’s risk and returns are

already spanned by the reference assets.

This seldom provides systematic evidence of

the existence of a listed infrastructure asset

class.



6. Results

Figure 3: Mean-variance frontier of 90% revenue threshold utilities and asset class referenceportfolio

●

●

●

●

●●

● Fixed Interest

Real Estate

Commodities

Hedge Funds

Developed Equities Emerging Equities

90% Utilities

−0.05

0.00

0.05

0.10

0.15

0.20


Ret

urns



An illustration of the findings in table 5 is

shown in figure 3, which presents themean-

variance frontier with and without the

addition of the naïve 90% Utilities portfolio

for the period January 2009 to December

2014. The results in table 5 confirm that this

portfolio does not improves the investment

opportunity set despite shifting the efficient

frontier to the left since the minimum

variance point is not statistically different

before and after adding ”infrastructure” to

the asset mix.

Next, we discuss our results using industry-

provided infrastructure indices as proxies

of the infrastructure asset class, testing

whether there are diversification benefits

with a global asset class-based reference

portfolio.

Global infrastructure indicesTable 6 presents our results for the global

infrastructure price and total return indices

described in section 5. Here, using price

returns for the full sample period (Panel

A, Table 6), the Huberman and Kandel

(1987) test finds an improvement in the

efficient frontier in six of the eight infras-

tructure indices examined. However, the

more restrictive Kan and Zhou (2012) test

only finds two of the eight global infras-

tructure indices to improve the reference

efficient frontier: the Dow Jones Brookfield

Global Infrastructure index and the UBS

Infrastructure and Utilities index.



6. Results

Other indices found to improve the efficient

frontier by the Huberman and Kandel (1987)

test are instead found to improve the

tangency portfolio or theminimum variance

portfolio but not both. As a result, it cannot

be accepted that these indices improve the

efficient frontier.

Using total returns (Panel D, table 6), again

six of the eight global indices reject the

null of the Huberman and Kandel (1987)

test. Only the FTSE Core index fails to span

when either the price or total returns are

employed, while the MSCI World Infras-

tructure is not spanned by the reference

asset classes using price returns but is when

considering total returns. The reverse is true

for the UBS 50-50.

Using the Kan and Zhou (2012) test, four

of the eight global infrastructure indices are

found to improve the efficient frontier. But

table 6 shows that most indices found by

the Huberman and Kandel (1987) test to

improve the efficient frontier only improved

the minimum variance portfolio, but not

improve the tangency portfolio. As a result,

it is not possible to conclude that these

listed infrastructure indices are not spanned

by existing asset classes.

Turning to sub-periods, for price returns

pre-GFC (Panels B and C of table 6),

the Huberman and Kandel (1987) test

finds that four of the eight global listed

infrastructure indices improve the efficient

frontier. However, the Kan and Zhou (2012)

test finds that these indices only improve

the minimum variance portfolio, and not

the tangency portfolio.

When total returns are considered for the

same period, the Huberman and Kandel

(1987) test finds all of the listed infras-

tructure indices improve the efficient

frontier. The results of the Kan and Zhou

(2012) test however, indicate that while

during this period the global indices

improved on the tangency portfolio, not all

impacted the minimum variance portfolio.

As a result, pre-GFC, only FTSE Core, MSCI

World Infrastructure and the MSCI ACWI

Capped infrastructure indices can be said

to improve the efficient frontier, as they

both improve the tangency portfolio and

reduce the minimum variance portfolio for

this period.

Post-GFC (panels E and F of table 6), pre-GFC

results are invalidated. Using price returns,

only one of the eight indices examined

is found to improve the efficient frontier

under both the Huberman and Kandel

(1987) and Kan and Zhou (2012) tests: the

Dow Jones Brookfield index. Using total

returns again only one index is found to

improve the efficient frontier under both

the Huberman and Kandel (1987) and Kan

and Zhou (2012) tests: the MSCI ACWI

Capped index.

Hence, pre-GFC results are not persistent

post-GFC. These results argue against the

existence of a well-defined and persistent

listed infrastructure ”asset class”.



6. Results

Figure 4: Illustration of the difference tests of mean-variance spanning of the FTSEMacquarie USA Infrastructure Index

(a) 2000-2014, asset class and factor-based reference

●●

●

●

●

●

●

●

●

Government Bonds

Corporate Bonds

High Yield

Hedge Funds

Commodities

Real Estate

US Equities



−0.05

0.00

0.05

0.10

0.15

0.20


Ret

urns



●

●

●

●

●

●

Market

Size

Value

Term

Default


−0.05

0.00

0.05

0.10

0.15

0.20


Ret

urns



(b) 2000-2008, asset class and factor-based reference

●

●

●

●

●

●

●

●

●

Government Bonds

Corporate Bonds

High Yield

Hedge Funds

Commodities

Real Estate

US Equities World Ex−US Equities


−0.30

−0.25

−0.20

−0.15

−0.10

−0.05

0.00

0.05

0.10

0.15

0.20

0.25

0.00 0.05 0.10 0.15 0.20 0.25 0.30 0.35 0.40Standard Deviation

Ret

urns



●

●

●

●

●

●

●

Europe Stocks

US Stocks

Size

Value

Term

Default

FTSE Macq

−0.05

0.00

0.05

0.10

0.15

0.20


Ret

urns



(c) 2009-2014, asset class and factor-based reference

●

●

●

●

●

●

●

●

●

Government Bonds

Corporate Bonds

High Yield

Hedge Funds

Commodities

Real Estate

US Equities



−0.05

0.00

0.05

0.10

0.15

0.20

0.25


Ret

urns



●

●

●

●

●

●

Europe Stocks

US Stocks

Size

Value

Term

FTSE Macq

−0.05

0.00

0.05

0.10

0.15

0.20


Ret

urns





6. Results

6.1.2 U.S. Infrastructure IndicesThe results for U.S. listed infrastructure

indices, are presented in table 7.

The Huberman and Kandel (1987) results

in table 7 indicate that for the full period

(Panel A and D) both the price returns and

total returns of the Alerian Infrastructure

MLP index improves on the efficient frontier.

None of the other infrastructure indices

reject the null hypothesis that the existing

asset class investments span the risk and

returns provided by the listed infrastructure

indices. Figure 4 provides an illustration.

When the Kan and Zhou (2012) test is

employed, the conclusion that the Alerian

Infrastructure MLP index improves the

investment opportunity set is reversed for

both the price and total return indices.

Whilst the Kan and Zhou (2012) finds that

the tangency portfolio has improved, it

does not reject the null hypothesis that the

global minimum variance has improved. As a

result, it is not possible to conclude that the

inclusion of the Alerian Infrastructure MLP

index improves the investment universe. As

the other infrastructure indices do not reject

the null hypothesis, the same conclusions

apply.

When pre- and post-GFC sub-samples are

considered (Panels B, C, E and F), the

conclusion that listed infrastructure assets

don’t improve the investment universe, is

still supported. For the first sub-period,

only the total returns of the Alerian

Infrastructure MLP index rejects the null

hypothesis of the Huberman and Kandel

(1987) test, as illustrated by figure 5.

When the Kan and Zhou (2012) test is

employed, none of the indices can reject

both steps of the test. It is not possible to

conclude that the inclusion of infrastructure

indices improves the mean-variance of

traditional asset classes in that period.

Using the second sub-sample period,

none of the indices, either using total

or price returns, are found to reject the

null hypothesis leading to the conclusion

that none of the indices represent an

improvement an investor’s diversification

opportunities.

6.1.3 PFI portfolioFinally, we report the ability of our PFI

portfolio to improve the mean-variance

efficiency of a diversified investor in the

United Kingdom in table 8. For the complete

sample, the price return series does not

provide diversification benefits. However,

total return results are found to improve

on the reference efficient frontier when

investing over the entire period, as figure 6

illustrates. The total returns PFI portfolio

passes both the Huberman and Kandel

(1987) and the Kan and Zhou (2012) tests

for the full sample period.

Looking at sub-periods in panels, diversifi-

cation benefits appear only in the period

following the GFC. Prior to the GFC, neither

price nor total returns of the PFI portfolios

improve the efficient frontier. Total returns

for example produce diversification benefits



6. Results

Figure 5: Mean-Variance Frontier of Alerian MLP Index Asset Class Proxies

●●

●

●

●

●

●

●

●

Government Bonds Corporate Bonds

High Yield

Hedge Funds

Commodities

Real Estate

US Equities


Alerian Infrastructure MLPs

−0.05

0.00

0.05

0.10

0.15

0.20

0.25


Ret

urns



Figure 6: Mean-Variance Frontier of total returns PFI portfolio and reference portfolio, 2000-2014

●

●

●

●

●

●

●

UK Gilts

Real Estate

Hedge Funds

Commodities

UK Equities

World Ex−UK Equities

PFI Portfolio

−0.05

0.00

0.05

0.10

0.15

0.20


Ret

urns





6. Results

according to the Huberman and Kandel

(1987) test, but the Kan and Zhou (2012)

test finds that these benefits are only due

to a change in the global minimum variance

portfolio. Without a corresponding increase

in the tangency portfolio, it is not possible

to conclude that the efficient frontier has

been improved. Still, these results may be

considered inconclusive, as PFI portfolio

returns begin in 2006.

After the GFC however, the price returns

of the PFI portfolio, pass the Huberman

and Kandel (1987) test, but the Kan and

Zhou (2012) finds that this is only due to

the improvement of the minimum variance

portfolio but not the tangency portfolio.

However, the total returns PFI portfolio is

found to exhibit diversification benefits by

both the Huberman and Kandel (1987) and

Kan and Zhou (2012) tests.

Hence, the impact of the PFI portfolio

appears to be one of the most persistent

of the various ”infrastructure” portfolios

that were tested on a total returns basis.

It improves diversification for the entire

investment period and, crucially, post-GFC,when all but one of the other infras-tructure indices fails to pass the post-GFC test of persistence.

6.2 Factor-based mean-variancespanning test resultsNext, we examine how the different listed

infrastructure definitions proposed above

fare against a factor-based reference

portfolio, i.e. whether investing in listed

infrastructure creates an exposure to

a combination of factors, which is not

otherwise available to investors already

allocating to the well-known factors

described in section 5.4.

As above we first present our results

for listed infrastructure companies

(section 6.2.1), followed by global listed

infrastructure (section 6.2.2) and US

(section 6.2.3) infrastructure indices. Unfor-

tunately, as this stage, we cannot build a

reference factor portfolio for the UK for

lack of sufficient data.

6.2.1 Listed infrastructure companiesTable 9 presents the our results for the

infrastructure portfolios using the naïve

infrastructure definition proposed in

section 5. Using the full sample (panels

A and D in table 9), the Huberman and

Kandel (1987) rejects the null hypothesis

that the efficient frontier is not improved

in five of the nine price return indices and

6 of the nine total return indices. Applying

the Kan and Zhou (2012) test however,

there is no evidence that infrastructure,

thus defined, provides diversification

benefits. Indices that qualified under the

Huberman and Kandel (1987) test all fail

to reject the null hypothesis for both

steps of the Kan and Zhou (2012) test.

Consistent with the findings for the asset

class reference portfolio, the addition

of listed infrastructure companies to a

factor-based allocation does not improve

the mean-variance frontier.



6. Results

Pre- and post-GFC results are consistent

with the full sample. In the period January

2000 to December 2008 (panels B and E

in table 9), eight of the nine price return

indices are found to improve the efficient

frontier according to the Huberman and

Kandel (1987) test. When the Kan and Zhou

(2012) test is applied however, this positive

result is overturned with none of the indices

examined passing the two-stage test. When

total returns are employed the results are

the same.

For the period January 2009 to December

2014 (panels C and F in table 9), results

mirror the pre-GFC sample. For the price

return indices the Huberman and Kandel

(1987) test finds that the mean variance

frontier is improved in six of the naïve

infrastructure portfolios. However, the Kan

and Zhou (2012) test results do not support

these findings and none of the portfolios

qualify. The total returns for naïve infras-

tructure portfolios result in the same

conclusions.

6.2.2 Global infrastructure indicesThe results for the spanning tests for global

listed infrastructure indices are presented in

table 10. The results are now familiar.

Using price returns for the full sample

(panels A in table 10), six of the eight indices

examined reject the null of the Huberman

and Kandel (1987) test at the 5% level,

but the Kan and Zhou (2012) test indicates

that only two of the eight indices improve

both the tangency portfolio as well as the

global minimum variance portfolio: only the

Dow Jones Brookfield and FTSE Core Infras-

tructure index can be said to improve the

reference efficient frontier.

For the period January 2000 to December

2008 (panels B in table 10), only four of the

eight indices are found to reject the null of

the Huberman and Kandel (1987) test at the

5% level, but none of these pass the Kan and

Zhou (2012) test. Between January 2009

and December 2014 (panels C in table 10)

only two of the eight portfolios are found to

reject the null of the Huberman and Kandel

(1987) test at the 5% level. Of these, only the

Dow Jones Brookfield is found to improve

the efficient frontier by the Kan and Zhou

(2012) test.

Using total returns for the full sample period

(panels D in table 10), all infrastructure

indices examined reject the null of the

Huberman and Kandel (1987) test at the 5%

level; and four still pass the Kan and Zhou

(2012) test: the S&P Global Infrastructure,

FTSE Macquarie Index, MSCI ACWI Capped

Index and the UBS 50-50 Index.

The same is true when the period January

2000 to December 2008 (panels E in

table 10), is considered: all indices pass

the Huberman and Kandel (1987) test and

three (The FTSE Core, MSCI World Infras-

tructure and MSCI ACWI Infrastructure) are

found by the Kan and Zhou (2012) test

to improve the tangency portfolio and the

global minimum variance portfolio, with The

remainder are found to only improve the

tangency portfolio.



6. Results

However, from January 2009 to December

2014, only three of the eight portfolios pass

the Huberman and Kandel (1987) test and

only one (the MSCI ACWI Capped Index) is

found to improve the efficient frontier by

the Kan and Zhou (2012) test on a total

return basis.

6.2.3 U.S. infrastructure indicesFinally, Table 11 shows the same results

using U.S. market indices and factors. For the

full sample period and using price returns

(Panel A in Table 11), the Alerian MLP Infras-

tructure index is, again, the only index found

to improve the efficient frontier using the

Huberman and Kandel (1987) test.

In the period from January 2000 to

December 2008 (Panel B in Table 11), the

Alerian MLP Infrastructure Index rejects

the null hypothesis of the Huberman and

Kandel (1987) test at the 5% level, but the

Kan and Zhou (2012) test concludes that

only the tangency portfolio has improved.

In the post-GFC period (Panel C in Table 11),

similar conclusions hold

Using total returns for the full sample period

(Panel D in Table 11), only the Alerian MLP

Infrastructure Index is again found to pass

the Huberman and Kandel (1987) test, but

the results of the Kan and Zhou (2012)

test indicate that this is due to the Alerian

MLP Infrastructure Index only improving the

tangency portfolio.

From January 2000 to December 2008

(Panel E in Table 11), conclusions are the

same. However, for the period from January

2009 to December 2014 (Panel F in Table 11),

the Alerian MLP index passes both the

Huberman and Kandel (1987) and Kan

and Zhou (2012) tests, indicating that the

efficient frontier has been improved.

Hence, the MLP index is found to have a

somewhat similar spanning profile than the

PFI portfolio in the sense that it manages

to create diversification benefits both before

after the GFC when considered from a total

return perspective.



6. Results

Table5:

Mea

n-va

rianc

esp

anning

resu

ltsfo

rNaïve

infras

truc

ture

portfo

lioswith

asse

tclass-b

uilt

referenc

epo

rtfo

lio

Pricereturns

,Pan

elA

50%

RevTe

l.50

%Re

vTran

sp.

50%

RevUtil.

75%

RevTe

l.75

%Re

vTran

sp.

75%

RevUtil.

90%

RevTe

l.90

%Re

vTran

sp.

90%

RevUtil.

H&

K5.

9995

1.43

991.

4685

4.43

469.

4460

1.16

294.

7566

0.11

400.

9837

p-va

lue

0.00

300.

2398

0.23

310.

0132

0.00

010.

3150

0.00

980.

8923

0.37

60St

epdo

wn1

11.4

124

0.66

730.

8408

8.16

720.

0605

0.37

878.

7251

0.18

940.

6453

p-va

lue

0.00

090.

4151

0.36

040.

0048

0.80

600.

5391

0.00

360.

6639

0.42

29St

epdo

wn2

0.55

342.

2168

2.09

810.

6742

18.9

336

1.95

410.

7545

0.03

881.

3248

p-va

lue

0.45

790.

1383

0.14

930.

4127

0.00

000.

1639

0.38

620.

8440

0.25

13,P

anel

B50

%Re

vTe

l.50

%Re

vTran

sp.

50%

RevUtil.

75%

RevTe

l.75

%Re

vTran

sp.

75%

RevUtil.

90%

RevTe

l.90

%Re

vTran

sp.

90%

RevUtil.

H&

K2.

7789

1.03

270.

3148

1.58

475.

1403

0.57

911.

3413

0.56

410.

2671

p-va

lue

0.06

690.

3598

0.73

070.

2101

0.00

750.

5622

0.26

620.

5707

0.76

61St

epdo

wn1

4.70

520.

5691

0.15

162.

6911

0.10

910.

3918

2.44

540.

0061

0.16

06p-

value

0.03

240.

4524

0.69

790.

1041

0.74

190.

5328

0.12

100.

9379

0.68

94St

epdo

wn2

0.82

251.

5027

0.48

200.

4704

10.2

620

0.77

120.

2339

1.13

330.

3768

p-va

lue

0.36

660.

2231

0.48

910.

4944

0.00

180.

3819

0.62

970.

2896

0.54

07,P

anel

C50

%Re

vTe

l.50

%Re

vTran

sp.

50%

RevUtil.

75%

RevTe

l.75

%Re

vTran

sp.

75%

RevUtil.

90%

RevTe

l.90

%Re

vTran

sp.

90%

RevUtil.

H&

K1.

9650

0.19

619.

4648

2.55

621.

5677

7.22

592.

4927

1.82

305.

5060

p-va

lue

0.14

840.

8224

0.00

020.

0854

0.21

630.

0015

0.09

060.

1697

0.00

62St

epdo

wn1

2.16

760.

3829

13.5

665

2.78

280.

2542

11.5

787

4.42

320.

0683

8.97

67p-

value

0.14

580.

5382

0.00

050.

1001

0.61

590.

0011

0.03

930.

7946

0.00

39St

epdo

wn2

1.73

190.

0094

4.50

522.

2683

2.91

422.

4763

0.53

443.

6290

1.81

58p-

value

0.19

270.

9229

0.03

750.

1368

0.09

250.

1204

0.46

730.

0611

0.18

24To

talretur

ns,P

anel

D50

%Re

vTe

l.50

%Re

vTran

sp.

50%

RevUtil.

75%

RevTe

l.75

%Re

vTran

sp.

75%

RevUtil.

90%

RevTe

l.90

%Re

vTran

sp.

90%

RevUtil.

H&

K4.

1227

1.09

440.

7730

2.83

729.

9398

0.70

092.

8997

0.85

470.

3654

p-va

lue

0.01

780.

3371

0.46

320.

0613

0.00

010.

4975

0.05

770.

4272

0.69

45St

epdo

wn1

7.81

850.

0279

0.00

695.

0241

0.36

570.

1568

4.82

711.

6359

0.05

90p-

value

0.00

580.

8677

0.93

400.

0263

0.54

620.

6926

0.02

930.

2026

0.80

83St

epdo

wn2

0.41

092.

1730

1.54

790.

6356

19.5

852

1.25

120.

9514

0.07

320.

6754

p-va

lue

0.52

230.

1423

0.21

510.

4264

0.00

000.

2649

0.33

070.

7870

0.41

23,P

anel

E50

%Re

vTe

l.50

%Re

vTran

sp.

50%

RevUtil.

75%

RevTe

l.75

%Re

vTran

sp.

75%

RevUtil.

90%

RevTe

l.90

%Re

vTran

sp.

90%

RevUtil.

H&

K2.

3508

0.61

690.

6647

1.22

374.

7298

1.15

150.

8439

1.13

200.

7195

p-va

lue

0.10

050.

5416

0.51

670.

2985

0.01

090.

3203

0.43

300.

3265

0.48

95St

epdo

wn1

3.79

870.

2847

0.87

811.

9978

0.02

201.

6295

1.61

510.

2724

1.15

84p-

value

0.05

410.

5948

0.35

100.

1606

0.88

240.

2047

0.20

670.

6029

0.28

44St

epdo

wn2

0.87

860.

9560

0.45

190.

4452

9.52

980.

6693

0.07

242.

0062

0.28

03p-

value

0.35

080.

3305

0.50

300.

5061

0.00

260.

4152

0.78

840.

1597

0.59

77,P

anel

F50

%Re

vTe

l.50

%Re

vTran

sp.

50%

RevUtil.

75%

RevTe

l.75

%Re

vTran

sp.

75%

RevUtil.

90%

RevTe

l.90

%Re

vTran

sp.

90%

RevUtil.

H&

K0.

4033

0.02

414.

7507

0.52

863.

0172

3.99

250.

2831

3.14

752.

9174

p-va

lue

0.66

980.

9762

0.01

190.

5919

0.05

580.

0232

0.75

440.

0496

0.06

12St

epdo

wn1

0.01

820.

0000

6.72

680.

0004

1.70

586.

8025

0.31

140.

4509

5.15

91p-

value

0.89

320.

9982

0.01

170.

9842

0.19

610.

0113

0.57

870.

5043

0.02

64St

epdo

wn2

0.80

040.

0490

2.55

311.

0731

4.28

281.

0868

0.25

745.

8931

0.63

57p-

value

0.37

420.

8255

0.11

490.

3040

0.04

240.

3010

0.61

360.

0179

0.42

81



6. Results

Table6:

Mea

n-va

rianc

esp

anning

test

resu

ltsfo

rgloba

llistedinfras

truc

ture

indice

swith

asse

tclass-b

uilt

referenc

epo

rtfo

lio

Pricereturns

,Pan

elA

Dow

Jone

sBr

ookfi

eld

S&PGloba

lInf

raFTSE

Mac

quarie

FTSE

Core

MSC

IWor

ldInfra

MSC

IACW

ICap

ped

UBS

UBS

50-5

0H

&K

6.23

7910.4

252

3.10

411.

1562

3.13

6416.4

068

7.84

833.

0324

p-va

lue

0.00

260.

0001

0.04

750.

3188

0.04

590.

0000

0.00

050.

0508

Step

down1

4.21

280.

0823

0.01

251.

2090

3.91

142.

6229

4.04

171.

5627

p-va

lue

0.04

200.

7746

0.91

100.

2742

0.04

950.

1072

0.04

590.

2130

Step

down2

8.07

4920.8

950

6.23

231.

1011

2.32

2429.9

118

11.4

546

4.48

76p-

value

0.00

520.

0000

0.01

350.

2965

0.12

930.

0000

0.00

090.

0356

,Pan

elB

Dow

Jone

sBr

ookfi

eld

S&PGloba

lInf

raFTSE

Mac

quarie

FTSE

Core

MSC

IWor

ldInfra

MSC

IACW

ICap

ped

UBS

UBS

50-5

0H

&K

3.01

648.

6243

4.41

542.

2605

2.93

7710.5

604

5.55

022.

5240

p-va

lue

0.05

600.

0004

0.01

470.

1230

0.05

760.

0001

0.00

520.

0852

Step

down1

1.80

740.

9489

1.08

742.

4140

0.73

022.

0287

3.22

963.

0950

p-va

lue

0.18

360.

3331

0.29

970.

1315

0.39

480.

1575

0.07

530.

0816

Step

down2

4.17

3516.3

104

7.73

632.

0090

5.15

8918.8

996

7.70

091.

9133

p-va

lue

0.04

510.

0001

0.00

650.

1670

0.02

520.

0000

0.00

660.

1696

,Pan

elC

Dow

Jone

sBr

ookfi

eld

S&PGloba

lInf

raFTSE

Mac

quarie

FTSE

Core

MSC

IWor

ldInfra

MSC

IACW

ICap

ped

UBS

UBS

50-5

0H

&K

3.75

961.

1947

1.28

670.

2095

0.80

073.

0735

2.13

291.

3365

p-va

lue

0.02

850.

3093

0.28

310.

8115

0.45

340.

0530

0.12

670.

2699

Step

down1

3.72

660.

8322

2.51

640.

4114

1.59

210.

4940

0.29

780.

5041

p-va

lue

0.05

790.

3650

0.11

750.

5235

0.21

150.

4847

0.58

720.

4802

Step

down2

3.64

211.

5613

0.05

570.

0078

0.00

925.

6966

4.01

062.

1854

p-va

lue

0.06

070.

2159

0.81

410.

9300

0.92

400.

0199

0.04

930.

1441

Totalretur

ns,P

anel

DDow

Jone

sBr

ookfi

eld

S&PGloba

lInf

raFTSE

Mac

quarie

FTSE

Core

MSC

IWor

ldInfra

MSC

IACW

ICap

ped

UBS

UBS

50-5

0H

&K

8.71

3813.8

824

4.58

562.

8989

2.02

5317.6

365

8.95

424.

8912

p-va

lue

0.00

030.

0000

0.01

150.

0597

0.13

510.

0000

0.00

020.

0086

Step

down1

7.92

552.

9554

0.88

983.

9974

0.80

278.

3858

7.93

784.

3504

p-va

lue

0.00

560.

0877

0.34

690.

0483

0.37

150.

0043

0.00

540.

0385

Step

down2

9.04

8024.4

921

8.28

681.

7490

3.25

1625.7

925

9.58

825.

3294

p-va

lue

0.00

310.

0000

0.00

450.

1890

0.07

310.

0000

0.00

230.

0221

,Pan

elE

Dow

Jone

sBr

ookfi

eld

S&PGloba

lInf

raFTSE

Mac

quarie

FTSE

Core

MSC

IWor

ldInfra

MSC

IACW

ICap

ped

UBS

UBS

50-5

0H

&K

4.10

1111.4

168

6.01

742.

9031

2.96

9611.4

718

6.27

683.

8680

p-va

lue

0.02

110.

0000

0.00

350.

0715

0.05

590.

0000

0.00

270.

0241

Step

down1

3.19

924.

0673

2.91

753.

3744

0.02

824.

7157

5.66

795.

4413

p-va

lue

0.07

840.

0472

0.09

090.

0769

0.86

690.

0323

0.01

920.

0217

Step

down2

4.83

9318.0

563

8.93

702.

2478

5.96

8417.5

811

6.58

152.

1979

p-va

lue

0.03

140.

0001

0.00

360.

1446

0.01

630.

0001

0.01

180.

1413

,Pan

elF

Dow

Jone

sBr

ookfi

eld

S&PGloba

lInf

raFTSE

Mac

quarie

FTSE

Core

MSC

IWor

ldInfra

MSC

IACW

ICap

ped

UBS

UBS

50-5

0H

&K

5.92

701.

3337

0.27

731.

1776

0.17

523.

7857

2.56

621.

2466

p-va

lue

0.00

430.

2706

0.75

870.

3145

0.83

970.

0278

0.08

460.

2943

Step

down1

7.15

690.

0031

0.51

852.

1403

0.01

433.

5961

0.93

950.

0490

p-va

lue

0.00

940.

9556

0.47

410.

1483

0.90

510.

0624

0.33

600.

8255

Step

down2

4.29

642.

7051

0.03

640.

2113

0.34

113.

8247

4.19

672.

4799

p-va

lue

0.04

210.

1048

0.84

930.

6473

0.56

120.

0547

0.04

450.

1201



6. Results

Table 7: Mean Variance Span test results for the U.S. listed infrastructure indices with assetclass-built reference portfolio

Price returns, Panel A Alerian MLP FTSE Macquarie USA MSCI USA Infrastructure MSCI USA Infra. CappedH & K 3.5537 0.9730 0.4046 1.7772p-value 0.0308 0.3821 0.6679 0.1722Stepdown1 6.8957 0.0604 0.0512 0.3592p-value 0.0094 0.8064 0.8212 0.5498Stepdown2 0.2048 1.9064 0.7622 3.2072p-value 0.6515 0.1709 0.3839 0.0751, Panel B Alerian MLP FTSE Macquarie USA MSCI USA Infrastructure MSCI USA Infra. CappedH & K 1.9041 0.3161 0.4258 2.2540p-value 0.1544 0.7349 0.6545 0.1104Stepdown1 3.7423 0.5859 0.0962 0.1576p-value 0.0559 0.4588 0.7570 0.6922Stepdown2 0.0640 0.0477 0.7623 4.3878p-value 0.8008 0.8305 0.3847 0.0388, Panel C Alerian MLP FTSE Macquarie USA MSCI USA Infrastructure MSCI USA Infra. CappedH & K 0.8072 0.1190 0.0891 0.0867p-value 0.4507 0.8880 0.9148 0.9171Stepdown1 1.6137 0.0577 0.1382 0.0350p-value 0.2086 0.8109 0.7114 0.8523Stepdown2 0.0007 0.1830 0.0407 0.1405p-value 0.9797 0.6702 0.8408 0.7090

Total returns, Panel D Alerian MLP FTSE Macquarie USA MSCI USA Infrastructure MSCI USA Infra. CappedH & K 6.0024 1.1463 0.6895 1.9286p-value 0.0030 0.3227 0.5032 0.1485Stepdown1 11.6155 0.0121 0.0246 0.9511p-value 0.0008 0.9128 0.8756 0.3308Stepdown2 0.3666 2.3070 1.3622 2.9069p-value 0.5456 0.1325 0.2448 0.0900, Panel E Alerian MLP FTSE Macquarie USA MSCI USA Infrastructure MSCI USA Infra. CappedH & K 3.8572 0.2963 0.5864 2.3219p-value 0.0244 0.7488 0.5583 0.1035Stepdown1 7.4502 0.4998 0.0063 0.2440p-value 0.0075 0.4931 0.9368 0.6225Stepdown2 0.2480 0.0966 1.1783 4.4336p-value 0.6196 0.7609 0.2803 0.0378, Panel F Alerian MLP FTSE Macquarie USA MSCI USA Infrastructure MSCI USA Infra. CappedH & K 0.7444 0.1307 0.4085 0.3702p-value 0.4791 0.8777 0.6664 0.6921Stepdown1 1.3711 0.0963 0.7868 0.3305p-value 0.2460 0.7574 0.3784 0.5674Stepdown2 0.1171 0.1674 0.0304 0.4143p-value 0.7333 0.6838 0.8622 0.5221

Table 8: Mean-variance spanning test results for the PFI stocks with asset class-builtreference portfolio

Price returns, Full SampleH & K 0.4211p-value 0.6575Stepdown1 0.0000p-value 0.9997Stepdown2 0.8508p-value 0.3586, Pre-GFCH & K 0.4173p-value 0.6630Stepdown1 0.6413p-value 0.4302Stepdown2 0.1959p-value 0.6615, Post-GFCH & K 7.9863p-value 0.0008Stepdown1 1.1721p-value 0.2830Stepdown2 14.7621p-value 0.0003

Total returns, Full SampleH & K 5.7016p-value 0.0045Stepdown1 5.4260p-value 0.0219Stepdown2 5.7239p-value 0.0186, Pre-GFCH & K 1.3958p-value 0.2649Stepdown1 1.4184p-value 0.2440Stepdown2 1.3529p-value 0.2546, Post-GFCH & K 7.6910p-value 0.0010Stepdown1 9.1745p-value 0.0035Stepdown2 5.5234p-value 0.0218



6. Results

Table9:

Mea

nVa

rianc

eSp

antest

resu

ltsfo

rthe

factor

asse

tclass

andth

eNaïve

infras

truc

ture

portfo

lioswith

factor

-bas

edreferenc

epo

rtfo

lios

Pricereturns

,FullS

ample

50%

RevTe

l.50

%Re

vTran

sp.

50%

RevUtil.

75%

RevTe

l.75

%Re

vTran

sp.

75%

RevUtil.

90%

RevTe

l.90

%Re

vTran

sp.

90%

RevUtil.

H&

K47.8

309

2.63

351.

8876

41.8

602

3.49

731.

3880

37.0

500

3.32

500.

9726

p-va

lue

0.00

000.

0747

0.15

450.

0000

0.03

240.

2523

0.00

000.

0383

0.38

01St

epdo

wn1

0.02

681.

0904

0.07

960.

0072

6.03

920.

1121

0.02

931.

8583

0.05

51p-

value

0.87

020.

2978

0.77

810.

9323

0.01

500.

7382

0.86

440.

1746

0.81

47St

epdo

wn2

96.1

669

4.17

443.

7149

84.1

880

0.92

892.

6775

74.4

815

4.76

851.

9003

p-va

lue

0.00

000.

0425

0.05

550.

0000

0.33

650.

1036

0.00

000.

0303

0.16

98,P

re-G

FC50

%Re

vTe

l.50

%Re

vTran

sp.

50%

RevUtil.

75%

RevTe

l.75

%Re

vTran

sp.

75%

RevUtil.

90%

RevTe

l.90

%Re

vTran

sp.

90%

RevUtil.

H&

K19.1

594

3.49

474.

5904

19.7

341

1.55

643.

1284

14.6

323

6.64

932.

6374

p-va

lue

0.00

000.

0340

0.01

230.

0000

0.21

590.

0480

0.00

000.

0019

0.07

64St

epdo

wn1

1.32

410.

1873

1.81

451.

1258

1.55

291.

4435

0.97

630.

3731

0.96

05p-

value

0.25

260.

6661

0.18

090.

2912

0.21

560.

2324

0.32

540.

5427

0.32

94St

epdo

wn2

36.8

786

6.85

627.

3084

38.2

956

1.55

154.

7927

28.2

949

13.0

045

4.31

60p-

value

0.00

000.

0102

0.00

800.

0000

0.21

570.

0308

0.00

000.

0005

0.04

02,P

ost-GFC

50%

RevTe

l.50

%Re

vTran

sp.

50%

RevUtil.

75%

RevTe

l.75

%Re

vTran

sp.

75%

RevUtil.

90%

RevTe

l.90

%Re

vTran

sp.

90%

RevUtil.

H&

K8.

5824

2.04

544.

8455

5.53

393.

2293

0.89

116.

7757

3.74

880.

7437

p-va

lue

0.00

050.

1373

0.01

080.

0060

0.04

580.

4150

0.00

210.

0286

0.47

92St

epdo

wn1

0.00

320.

2274

5.72

120.

3092

3.38

581.

7022

0.03

401.

2738

1.18

17p-

value

0.95

480.

6350

0.01

960.

5800

0.07

020.

1965

0.85

420.

2631

0.28

09St

epdo

wn2

17.4

170

3.90

783.

7121

10.8

691

2.96

870.

0791

13.7

121

6.19

880.

3049

p-va

lue

0.00

010.

0521

0.05

820.

0016

0.08

940.

7794

0.00

040.

0152

0.58

26To

talretur

ns,F

ullS

ample

50%

RevTe

l.50

%Re

vTran

sp.

50%

RevUtil.

75%

RevTe

l.75

%Re

vTran

sp.

75%

RevUtil.

90%

RevTe

l.90

%Re

vTran

sp.

90%

RevUtil.

H&

K49.0

714

4.14

652.

4766

44.8

485

5.33

902.

0835

38.3

526

4.70

651.

6385

p-va

lue

0.00

000.

0174

0.08

700.

0000

0.00

560.

1276

0.00

000.

0102

0.19

72St

epdo

wn1

0.64

541.

4520

0.31

630.

6925

9.57

430.

8237

0.51

964.

0237

0.67

81p-

value

0.42

280.

2298

0.57

450.

4064

0.00

230.

3654

0.47

200.

0464

0.41

14St

epdo

wn2

97.6

941

6.82

354.

6550

89.1

603

1.05

243.

3467

76.3

942

5.29

822.

6038

p-va

lue

0.00

000.

0098

0.03

230.

0000

0.30

640.

0690

0.00

000.

0225

0.10

84,P

re-G

FC50

%Re

vTe

l.50

%Re

vTran

sp.

50%

RevUtil.

75%

RevTe

l.75

%Re

vTran

sp.

75%

RevUtil.

90%

RevTe

l.90

%Re

vTran

sp.

90%

RevUtil.

H&

K23.2

175

4.48

585.

2056

23.3

121

2.03

853.

7743

18.4

482

6.88

873.

3058

p-va

lue

0.00

000.

0136

0.00

700.

0000

0.13

550.

0262

0.00

000.

0016

0.04

06St

epdo

wn1

2.82

190.

1478

2.08

632.

2061

2.55

222.

3533

2.10

820.

9129

1.76

75p-

value

0.09

600.

7014

0.15

170.

1405

0.11

320.

1281

0.14

960.

3416

0.18

67St

epdo

wn2

42.8

549

8.89

748.

2379

43.9

038

1.50

235.

1278

34.4

179

12.8

754

4.80

83p-

value

0.00

000.

0036

0.00

500.

0000

0.22

310.

0256

0.00

000.

0005

0.03

06,P

ost-GFC

50%

RevTe

l.50

%Re

vTran

sp.

50%

RevUtil.

75%

RevTe

l.75

%Re

vTran

sp.

75%

RevUtil.

90%

RevTe

l.90

%Re

vTran

sp.

90%

RevUtil.

H&

K5.

2249

1.65

363.

3435

3.86

234.

5354

0.41

923.

7247

5.22

980.

3791

p-va

lue

0.00

780.

1991

0.04

130.

0259

0.01

420.

6593

0.02

930.

0078

0.68

59St

epdo

wn1

0.68

780.

9210

3.24

200.

4323

5.71

510.

7647

0.36

163.

4145

0.44

85p-

value

0.40

990.

3407

0.07

630.

5131

0.01

960.

3850

0.54

960.

0690

0.50

53St

epdo

wn2

9.80

712.

3890

3.33

527.

3537

3.13

810.

0739

7.15

506.

8036

0.31

23p-

value

0.00

260.

1268

0.07

220.

0085

0.08

100.

7866

0.00

940.

0112

0.57

81



6. Results

Table10

:Mea

nVa

rianc

eSp

antest

resu

ltsfo

rthe

Globa

llistedinfras

truc

ture

indice

swith

factor

-bas

edreferenc

epo

rtfo

lios

Pricereturns

,FullS

ample

Dow

Jone

sBr

ookfi

eld

S&PGloba

lInf

raFTSE

Mac

quarie

FTSE

Core

MSC

IWor

ldInfra

MSC

IACW

ICap

ped

UBS

UBS

50-5

0H

&K

6.23

7910.4

252

3.10

411.

1562

3.13

6416.4

068

7.84

833.

0324

p-va

lue

0.00

260.

0001

0.04

750.

3188

0.04

590.

0000

0.00

050.

0508

Step

down1

4.21

280.

0823

0.01

251.

2090

3.91

142.

6229

4.04

171.

5627

p-va

lue

0.04

200.

7746

0.91

100.

2742

0.04

950.

1072

0.04

590.

2130

Step

down2

8.07

4920.8

950

6.23

231.

1011

2.32

2429.9

118

11.4

546

4.48

76p-

value

0.00

520.

0000

0.01

350.

2965

0.12

930.

0000

0.00

090.

0356

,Pre-G

FCDow

Jone

sBr

ookfi

eld

S&PGloba

lInf

raFTSE

Mac

quarie

FTSE

Core

MSC

IWor

ldInfra

MSC

IACW

ICap

ped

UBS

UBS

50-5

0H

&K

3.01

648.

6243

4.41

542.

2605

2.93

7710.5

604

5.55

022.

5240

p-va

lue

0.05

600.

0004

0.01

470.

1230

0.05

760.

0001

0.00

520.

0852

Step

down1

1.80

740.

9489

1.08

742.

4140

0.73

022.

0287

3.22

963.

0950

p-va

lue

0.18

360.

3331

0.29

970.

1315

0.39

480.

1575

0.07

530.

0816

Step

down2

4.17

3516.3

104

7.73

632.

0090

5.15

8918.8

996

7.70

091.

9133

p-va

lue

0.04

510.

0001

0.00

650.

1670

0.02

520.

0000

0.00

660.

1696

,Pos

t-GFC

ADow

Jone

sBr

ookfi

eld

S&PGloba

lInf

raFTSE

Mac

quarie

FTSE

Core

MSC

IWor

ldInfra

MSC

IACW

ICap

ped

UBS

UBS

50-5

0H

&K

6.51

360.

6060

0.03

441.

8708

2.83

724.

4366

2.16

741.

7735

p-va

lue

0.00

260.

5485

0.96

620.

1620

0.06

570.

0155

0.12

240.

1776

Step

down1

8.37

450.

9974

0.05

952.

3981

0.50

682.

1816

4.31

893.

4760

p-va

lue

0.00

510.

3215

0.80

810.

1262

0.47

900.

1444

0.04

150.

0666

Step

down2

4.19

750.

2146

0.00

961.

3164

5.20

536.

5773

0.01

510.

0684

p-va

lue

0.04

430.

6447

0.92

240.

2553

0.02

570.

0125

0.90

250.

7944

Totalretur

ns,F

ullS

ample

Dow

Jone

sBr

ookfi

eld

S&PGloba

lInf

raFTSE

Mac

quarie

FTSE

Core

MSC

IWor

ldInfra

MSC

IACW

ICap

ped

UBS

UBS

50-5

0H

&K

9.35

1912.1

028

5.74

005.

1611

12.4

659

9.66

883.

9037

10.4

645

p-va

lue

0.00

020.

0000

0.00

390.

0073

0.00

000.

0001

0.02

200.

0001

Step

down1

18.0

756

15.6

201

6.11

677.

4542

1.88

609.

5355

7.78

6010.9

591

p-va

lue

0.00

000.

0001

0.01

440.

0075

0.17

140.

0023

0.00

590.

0011

Step

down2

0.55

907.

8277

5.20

472.

6972

22.9

290

9.34

380.

0207

9.43

02p-

value

0.45

590.

0058

0.02

380.

1036

0.00

000.

0026

0.88

570.

0025

,Pre-G

FCDow

Jone

sBr

ookfi

eld

S&PGloba

lInf

raFTSE

Mac

quarie

FTSE

Core

MSC

IWor

ldInfra

MSC

IACW

ICap

ped

UBS

UBS

50-5

0H

&K

7.04

868.

3514

4.57

949.

2948

21.4

088

7.45

604.

0881

8.32

79p-

value

0.00

170.

0005

0.01

260.

0007

0.00

000.

0010

0.01

960.

0004

Step

down1

13.2

588

16.6

604

8.70

279.

1231

4.48

439.

8566

8.00

8114.1

227

p-va

lue

0.00

050.

0001

0.00

400.

0051

0.03

660.

0022

0.00

560.

0003

Step

down2

0.70

870.

0354

0.42

257.

5010

37.0

790

4.65

520.

1573

2.24

68p-

value

0.40

290.

8512

0.51

720.

0101

0.00

000.

0333

0.69

250.

1370

,Pos

t-GFC

Dow

Jone

sBr

ookfi

eld

S&PGloba

lInf

raFTSE

Mac

quarie

FTSE

Core

MSC

IWor

ldInfra

MSC

IACW

ICap

ped

UBS

UBS

50-5

0H

&K

7.36

911.

2262

0.34

802.

6625

3.03

815.

9643

3.51

652.

3387

p-va

lue

0.00

130.

2999

0.70

740.

0771

0.05

460.

0041

0.03

530.

1043

Step

down1

11.0

624

2.07

310.

6576

4.24

552.

3362

6.92

646.

9994

4.38

85p-

value

0.00

140.

1546

0.42

030.

0432

0.13

110.

0105

0.01

020.

0400

Step

down2

3.20

200.

3734

0.03

861.

0304

3.66

794.

6012

0.03

090.

2753

p-va

lue

0.07

800.

5432

0.84

480.

3137

0.05

970.

0355

0.86

100.

6015



6. Results

Table 11: Mean Variance Span test results for the factor asset classes and U.S. listed infras-tructure indices with factor-based reference portfolios

Price returns, Full Sample Alerian MLP FTSE Macquarie USA MSCI USA Infrastructure MSCI USA Infra. CappedH & K 8.0362 1.0298 1.0553 0.1836p-value 0.0005 0.3613 0.3503 0.8324Stepdown1 8.7670 0.0449 0.5795 0.2557p-value 0.0035 0.8326 0.4475 0.6137Stepdown2 6.9949 2.0365 1.5348 0.1120p-value 0.0089 0.1571 0.2170 0.7383, Pre-GFC Alerian MLP FTSE Macquarie USA MSCI USA Infrastructure MSCI USA Infra. CappedH & K 3.2144 1.3422 0.0637 0.6650p-value 0.0443 0.2909 0.9383 0.5165Stepdown1 6.4173 0.0560 0.1059 0.6719p-value 0.0128 0.8161 0.7455 0.4143Stepdown2 0.0109 2.7932 0.0216 0.6601p-value 0.9171 0.1141 0.8833 0.4184, Post-GFC Alerian MLP FTSE Macquarie USA MSCI USA Infrastructure MSCI USA Infra. CappedH & K 0.8072 0.1190 0.0891 0.0867p-value 0.4507 0.8880 0.9148 0.9171Stepdown1 1.6137 0.0577 0.1382 0.0350p-value 0.2086 0.8109 0.7114 0.8523Stepdown2 0.0007 0.1830 0.0407 0.1405p-value 0.9797 0.6702 0.8408 0.7090

Total returns, Full Sample Alerian MLP FTSE Macquarie USA MSCI USA Infrastructure MSCI USA Infra. CappedH & K 12.4051 0.9491 1.3303 0.1821p-value 0.0000 0.3910 0.2671 0.8337Stepdown1 21.9107 0.1018 0.3802 0.3176p-value 0.0000 0.7505 0.5383 0.5738Stepdown2 2.5900 1.8147 2.2886 0.0468p-value 0.1093 0.1814 0.1321 0.8289, Pre-GFC Alerian MLP FTSE Macquarie USA MSCI USA Infrastructure MSCI USA Infra. CappedH & K 5.4190 0.9820 0.0060 0.5267p-value 0.0058 0.3974 0.9940 0.5921Stepdown1 9.8783 0.0571 0.0001 0.0966p-value 0.0022 0.8144 0.9938 0.7565Stepdown2 0.8830 2.0263 0.0121 0.9654p-value 0.3496 0.1738 0.9127 0.3282, Post-GFC Alerian MLP FTSE Macquarie USA MSCI USA Infrastructure MSCI USA Infra. CappedH & K 16.3809 0.8787 1.0270 1.6919p-value 0.0000 0.4201 0.3637 0.1921Stepdown1 8.4217 1.2882 0.8764 0.9380p-value 0.0050 0.2605 0.3526 0.3363Stepdown2 21.9127 0.4672 1.1798 2.4481p-value 0.0000 0.4966 0.2813 0.1224


7. Conclusions



7. Conclusions

7.1 Summary of resultsIn this paper, we examined the contention

that focusing on ”listed infrastructure”

has the potential to create diversification

benefits previously unavailable to large

investors already active in public markets.

The reasons for doing so were threefold:

1. Several papers have argued that it is the

case but do not provide robust statistical

tests of this hypothesis;




invest in listed infrastructure arguing

that it is an asset class in its own right;

3. Capital market instruments are often


in privately-held (unlisted) infrastructure

but the adequacy of such proxies remains

untested.

We tested the notion that there is a unique

and persistent ”listed infrastructure effect”

using 22 listed infrastructure proxies and a

series of statistical tests of mean-variance

spanning against reference portfolios, built

with either traditional asset classes or

investment factors. We conducted these

tests for global, U.S. and UK markets

covering the past 15 years, on a price return

and total return basis.

We conclude that listed infrastructure,as it is traditionally defined by SIC code

and industrial sector, is not an asset classor a unique combination of marketfactors, but instead cannot be persistently

distinguished from existing exposures

in investors’ portfolios. Expecting the

emergence of a new or unique ”infras-

tructure asset class” by focusing on public

equities selected on the basis of industrial

sectors is thus misguided. Such asset

selection schemes do not create diver-

sification benefits. Figure 4 provides an

illustration of these results in the case of

the FTSE Macquarie Listed Infrastructure

Index for the U.S. market.

Our test result are summarised in table 12.

Stylised facts include:

1. We tested 22 proxies of listed infras-

tructure and found little to no robust

evidence of a ”listed infrastructure asset

class” that was not already spanned by

a combination of capital market instru-

ments and alternatives or a factor-based

asset allocation;

2. The majority of test portfolios that

improved the mean-variance efficient

frontier before the GFC fail to repeat this

feat post GFC. There is no evidence of

persistent diversification benefits;

3. Of the 22 test portfolios used in this

paper to try and establish the existence

of a listed infrastructure asset class, only

four manage to improve on a typical

asset allocation defined either by tradi-

tional asset class or by factor exposure

after the GFC and only one is not spanned

both pre- and post-GFC. We return to

these in the discussion below;

4. Building baskets of stocks on the basis

of their SIC code and proportion of

infrastructure income fails generate a

convincing exposure to a new asset class.



7. Conclusions

Table 12: Summary table of mean-variance spanning tests: This table summarises thefindings of the mean-variance spanning tests for the infrastructure proxies and differentasset allocation strategies employed in this paper. ✓indicated that the infrastructure proxypassed all three spanning tests at the 5% confidence level either with reference to anasset class-build portfolio or a factor-built portfolio. The first column reports results for thewhole sample from 2000 to 2014, the next two columns report pre- and post-GFC resultsand the fourth column highlights the proxies that show post-GFC persistence of pre-GFCimprovement of the efficient portfolio frontier.

Full sample Pre-GFC Post-GFCPrice Returns50% Rev. Req. Telecom. × × ×50% Rev. Req. Transport × × ×50% Rev. Req. Utilities × × ×75% Rev. Req. Telecom. × × ×75% Rev. Req. Transport × × ×75% Rev. Req. Utilities × × ×90% Rev. Req. Telecom. × × ×90% Rev. Req. Transport × × ×90% Rev. Req. Utilities × × ×Alerian MLP ✓ × ×FTSE Macquarie USA × × ×MSCI USA × × ×MSCI USA Infra Capped × × ×DJ Brookfield Global ✓ × ✓S&P Infrastructure × ✓ ×FTSE Macquarie Infra × × ×FTSE Global Core × × ×MSCI World Infra × ✓ ×MSCI ACWI Infra Capped × × ×UBS Global Infra Uti ✓ ✓ ×UBS Global 50-50 × × ×PFI Portfolio × × ×Total Returns50% Rev. Req. Telecom. × × ×50% Rev. Req. Transport × × ×50% Rev. Req. Utilities × × ×75% Rev. Req. Telecom. × × ×75% Rev. Req. Transport × × ×75% Rev. Req. Utilities × × ×90% Rev. Req. Telecom. × × ×90% Rev. Req. Transport ✓ × ×90% Rev. Req. Utilities × × ×Alerian MLP × × ✓FTSE Macquarie USA × × ×MSCI USA × × ×MSCI USA Infra Capped × × ×DJ Brookfield Global ✓ × ✓FTSE Global Core × ✓ ×S&P Infrastructure ✓ ✓ ×FTSE Macquarie Infra ✓ ✓ ×MSCI World Infra × ✓ ×MSCI ACWI Infra Capped ✓ ✓ ✓UBS Global Infra Uti ✓ ✓ ×UBS Global 50-50 ✓ × ×PFI Portfolio ✓ × ✓



7. Conclusions

A more promising avenue is to focus

on underlying contractual or gover-

nance structures that tend to maximise

dividend payout and pay dividends with

great regularity, such as the PFI or MLP

models;

5. More generally, benchmarking unlisted

infrastructure investments with thematic

(industry-based) stock indices is unlikely

to be very helpful from a pure asset

allocation perspective i.e. the latter do

not exhibit a risk/return trade-off or

betas that large investors did not have

access to already.

7.2 DiscussionWhile we conclude from testing the impact

of 22 proxies, that there is no convincing

evidence of a listed infrastructure asset

class, it is worthwhile examining the four

proxies that manage to improve on the

proposed reference asset allocation afterthe GFC.

Indeed, high pre-GFC Sharpe ratios that do

not survive the 2008 credit crunch and lose

all statistical significance in mean-variance

spanning tests post-GFC do not make good

candidates for an asset class or bundle

of factors. However, proxies that pass the

mean-variance tests since 2008may at least

open the possibility of a more persistent

effect.

The four proxies that are not already

spanned by our reference portfolios in the

post GFC period questions are:

1. The Brookfield Dow Jones Infrastructure

Index: on close examination, this index

made a significant shift towards the

oil and gas sector after the GFC and

benefited from the significant rise of oil

prices in the subsequent period. We note,

without further investigation, that since

2014 and the collapse of global oil prices,

it has experienced lacklustre perfor-

mance. Hence, rather than an ”infras-

tructure effect”, this proxy may have

been capturing a kind of ”oil play”;

2. The MSCI ACWI Infrastructure Capped:

this proxy is the only one which passes

the spanning tests both pre- and post-

GFC. It is one of the few listed infras-

tructure indices which is not simply

weighted by market capitalisation but is

instead constrained to have a maximum

of one third of its assets invested in

Telecoms, one third in Utilities and

another third in energy and trans-

portation. Hence, it uses a very adhoc weighing scheme, vaguely resem-

bling equal weighting, which never-

theless improves on the market cap-

weighted point of reference. Again,

rather than an effect driven by a

hypothetical ”infrastructure asset class”,

it seems reasonable to assume that

portfolio weights explain the impact of

this proxy; 1010 - In future research, similar testof mean-variance spanning againstefficient or ”smart” reference indicesis necessary to control for sucheffects.

3. The Alerian MLP Index: this proxy and

the next one only improve the reference

allocation post GFC in a total return

basis. Here the role played by dividend

payouts, their size and regularity relative

to other stocks are likely candidates to

explain why they succeed in passing the



7. Conclusions

spanning tests. However, this index also

proves to be high risk and correlated with

the energy price cycle

4. The PFI Portfolio, because it corresponds

to self-contained investment vehicles

that receive a steady income stream

from the public sector and have risky

but predictable operating and financing

costs, and are, by design, likely to have

very regular dividend payouts.

This last point is important since the

observed improvement of the efficient

frontier by adding assets such as MLPs or

PFIs also corresponds to the beginning of

the very low interest rate policies introduced

by U.S. and U.K. central banks after the GFC.

In such an environment, such high-coupon

paying assets start to exhibit previously

unremarkable characteristics that, mechan-

ically increases their ability to have an

impact on the reference portfolio.

Crucially, what determines this ability to

deliver regular and high dividend payouts

is the contractual and governance structure

of the underlying businesses, not their

belonging to a given industrial sector, which

does not suggest any particular a prioridividend paying behaviour.

However, it must be noted that the relatively

low aggregate market capitalisation of

listed entities offering a ”clean” exposure to

infrastructure ”business models” as opposed

to infrastructure industrial sectors may limit

the ability of investors to enjoy these

potential benefits unless the far larger

unlisted infrastructure fund universe has

similar characteristics.

We conclude that as an asset selectionscheme, the notion of investing in ”infras-

tructure” (listed or not) should be under-

stood as a heuristic i.e. a mental short-

cut meant to create an exposure to certain

factors, but neither a thing nor an end in

itself.

A clear distinction can be made between

infrastructure as a matter of public policy,

in which case the focus is rightly on indus-

trial functions, and the point of view of

financial investors, who may be exposed

to completely different risks through

investments in firms providing exactly

the same industrial functions. Notional

grouping of assets by industrial sectors

(transport, energy, water, etc) create very

little information or predictive power.

Focusing on definitions of infrastructure

investment that match the tangible or

industrial characteristics of certain firms

or assets is unhelpful because it does not

take in to account the mechanisms that

create the potentially desirable character-

istics of infrastructure investment. Infras-

tructure investment should be construed

solely as a way to buy claims on future

cash flows created by specific underlying

business models, themselves the product

of long-term contractual arrangements

between public and private parties (or alter-

natively between two private parties).


8. Appendices



8. Appendices

Table 13: Descriptive statistics of annualised price and total returns of reference asset classes,2000-2014

Panel A: Descriptive statistics of annualised price and total returns of the global reference asset classes, 2000-2014

Price Returns, Bonds Real Estate Commo Hedge Funds OECD Stocks EM StocksPrice return 0.0545 0.0152 0.0224 0.0615 −0.0010 0.0162Risk 0.0584 0.1998 0.2355 0.0568 0.1611 0.2339SR 0.8445 0.0506 0.0734 0.9900 −0.0371 0.0478

Total Returns, Bonds Real Estate Commo Hedge Funds OECD Stocks EM StocksTot. return 0.0545 0.0543 −0.0193 0.0634 0.0227 0.0436Risk 0.0584 0.1993 0.2389 0.0567 0.1611 0.2336SR 0.8446 0.2462 −0.1014 1.0253 0.1093 0.1645

Panel B: Descriptive statistics of annualised price and total returns of U.S. reference assets classes, 2000-2014

Price Returns, Gov Bonds Corp. Bonds High Yld Real Estate Commo. H. Funds U.S. Stocks World

ex-U.S.Price return 0.0132 0.0109 −0.0062 0.0634 0.0104 0.0645 0.0293 0.0038Risk 0.0412 0.0344 0.0984 0.0567 0.2340 0.2186 0.1565 0.1739SR 0.1973 0.1701 −0.1130 1.0255 0.0227 0.2710 0.1546 −0.0072

Total Returns, Gov Bonds Corp. Bonds High Yld Real Estate Commo. H.Funds U.S. Stocks World

ex-U.S.Tot. return 0.0545 0.0578 0.0806 0.0653 0.0104 0.1192 0.0482 0.0325Risk 0.0419 0.0350 0.0996 0.0566 0.2340 0.2195 0.1566 0.1743SR 1.1778 1.4997 0.7554 1.0597 0.0227 0.5179 0.2743 0.1568

Panel C: Descriptive statistics of annualised price and total returns of the U.K. reference asset classes, 2000-2014

Price Returns, Fixed Interest Real Estate Commo. H. Funds UK. Stocks World ex-UKPrice return 0.0590 0.0305 0.0487 0.0126 −0.0036 0.0201Risk 0.0500 0.2074 0.1206 0.2171 0.1420 0.1548SR 1.0755 0.1224 0.3605 0.0347 −0.0603 0.0971

Total Returns, Fixed Interest Real Estate Commo. H. Funds UK. Stocks World ex-UKTot. return 0.0590 0.0657 0.0506 0.0126 0.0309 0.0429Risk 0.0498 0.2084 0.1206 0.2171 0.1423 0.1553SR 1.0778 0.2898 0.3759 0.0347 0.1809 0.2427

Table 14: Descriptive statistics of annualised price and total returns of the reference factors,2000-2014

Panel A: Descriptive statistics of annualised price and total returns of the global reference factors, 2000-2014

Price Returns, Eur Market U.S. Market Size ValuePrice return −0.0344 −0.0105 0.0522 0.0250Risk 0.1963 0.1562 0.0713 0.0949SR −0.1997 −0.0989 0.6587 0.2099

Total Returns, Eur Market U.S. Market Size Value Term DefaultTot. return −0.0031 0.0084 0.0467 0.0421 0.0387 −0.0069Risk 0.1962 0.1561 0.0717 0.0946 0.0758 0.2274SR −0.0413 0.0219 0.5786 0.3904 0.4422 −0.0522

Panel B: Descriptive statistics of annualised price and total returns of the U.S. reference factors, 2000-2014

Price Returns, Market Size Value Term DefaultPrice return 0.0162 0.0271 0.0290 −0.0022 −0.0242Risk 0.1587 0.1058 0.1143 0.0870 0.0680SR 0.0698 0.2081 0.2089 −0.0819 −0.4284

Total Returns, Market Size Value Term DefaultTot. return 0.0348 0.0216 0.0430 0.0551 0.0330Risk 0.1586 0.1061 0.1144 0.0871 0.0679SR 0.1867 0.1558 0.3305 0.5723 0.4109



8. Appendices

Table 15: List of project finance SPVs in the listed PFI portfolio

Project name Sector Investor Country Revenue SourceA249 Road Roads HICL UK Unitary ChargeA92 Road Roads HICL UK Unitary ChargeAbbotsford Hospital Hospitals JLIF Canada Unitary ChargeAbingdon -Thames ValleyPolice

Gov Services INPP UK Unitary Charge

Addiewell Prison Gov Services HICL UK Unitary ChargeAlberta Schools Gov Services INPP Canada Unitary ChargeAllenby and Connaught PFIProject, UK

Gov Services HICL UK Unitary Charge

Angel Trains Hospitals INPP UK Unitary ChargeAquasure Victorian Desali-nation Project, Australia

Gov Services HICL Australia Purchase agreement

Avon and Somerset Courts Gov Services JLIF UK Unitary ChargeBarking and Dagenham PFISPV 1

education INPP UK Unitary Charge

Barking and Havering Clinics Hospitals BBGI UK Unitary ChargeBarking and DagenhamSchools

education HICL UK Unitary Charge

Barnet and Haringey Clinics Hospitals BBGI UK Unitary ChargeBarnet Hospital, UK Hospitals HICL UK Unitary ChargeBarnet Street Lighting Gov Services JLIF UK Unitary ChargeBarnsley PFI SPV 1 education INPP, JLIF UK Unitary ChargeBarnsley PFI SPV 2 education INPP UK Unitary ChargeBarnsley PFI SPV 3 education INPP UK Unitary ChargeBBG Lakeside Hospitals INPP UK Unitary ChargeBedford Schools education BBGI UK Unitary ChargeBeNEX Rail Link INPP Germany unknownBentilee Hub CommunityCentre

Gov Services JLIF UK Unitary Charge

Bexley Schools education JLIF UK Unitary ChargeBexley, Bromley, Greenwich 1 Hospitals INPP UK Unitary ChargeBexley, Bromley, Greenwich 2 Hospitals INPP UK Unitary ChargeBHH Mt Vernon Hospitals INPP UK Unitary ChargeBHH Sudhury Hospitals INPP UK Unitary ChargeBirmingham and Solihull LIFT Hospitals HICL UK Unitary ChargeBirmingham Hospitals Hospitals HICL UK Unitary ChargeBirmingham PFI SPV 1 education INPP UK Unitary ChargeBishop Auckland Hospital, UK Hospitals HICL UK Unitary ChargeBistol PFI SPV 1 education INPP UK Unitary ChargeBlackburn PFI SPV 1 education INPP, HICL UK Unitary ChargeBlackburn PFI SPV 2 education INPP UK Unitary ChargeBlackpool Primary Care Facility Hospitals HICL UK Unitary ChargeBMBF education INPP Germany Unitary ChargeBoldon School education HICL UK Unitary ChargeBradford BSF Phase 2 education HICL,INPP UK Unitary ChargeBradford PFI SPV 1 education INPP UK Unitary ChargeBrent, Harrow, Hillingdon Hospitals INPP UK Unitary ChargeBrentwood CommunityHospital

Hospitals HICL UK Unitary Charge

Brescia Hospital Hospitals INPP Portugal Unitary ChargeBrighton Hospital, UK Hospitals HICL UK Unitary ChargeBristol BSF education JLIF UK Unitary ChargeBristol Fishponds andHampton House

Hospitals INPP UK Unitary Charge

Bristol Shirehampton andWhitchurch


Brockley Social Housing PFI Gov Services JLIF UK Unitary ChargeBurg Prison Gov Services BBGI Germany Unitary ChargeCalderdale education INPP UK Unitary ChargeCambridgeshire PFI SPV 1 education INPP UK Unitary ChargeCamden Housing Gov Services JLIF UK Unitary ChargeCanning Town Social HousingPFI


Central Middlesex Hospital, UK Hospitals HICL UK Unitary ChargeClackmannanshire Schools education BBGI UK Unitary ChargeCleveland Police Station andHQ


Connect PFI Roads HICL UK Unitary ChargeConwy Schools, UK education HICL UK Unitary ChargeCork School of Music education HICL Ireland Unitary ChargeCoventry Schools education BBGI UK Unitary ChargeCroydon Schools education HICL UK Unitary ChargeDarlington Schools, UK education HICL UK Unitary ChargeDefence Sixth Form College,UK


Derby City PFI SPV 1 education INPP UK Unitary ChargeDerby Courts Gov Services INPP UK Unitary ChargeDerby Schools education INPP, HICL UK Unitary ChargeDerby Schools 2 education INPP UK Unitary ChargeDerbyshire PFI SPV 1 education INPP UK Unitary ChargeDiabolo (T2 and T3 and T5) Rail Link INPP UK unknown

source: annual reports



8. Appendices

Table 16: List of project finance SPVs in the listed PFI portfolio (continued)

Project name Sector Investor Country Revenue SourceDoncaster Mental Health Hospitals HICL UK Unitary ChargeDoncaster Schools education HICL UK Unitary ChargeDorset Fire and Rescue Gov Services HICL UK Unitary ChargeDublin Courts Gov Services INPP Ireland Unitary ChargeDudley Brierly Hill Hospitals INPP UK Unitary ChargeDudley Ridge Hill and Stour-bridge


Durham and Cleveland PoliceTactical Training Centre


Durham Courts Gov Services INPP Canada Unitary ChargeDurham PFI SPV 1 education INPP UK Unitary ChargeDutch High Speed Rail Link Roads HICL Netherlands Unitary ChargeE18 Motorway Roads BBGI Norway Unitary ChargeE18 Road Roads JLIF Finland Unitary ChargeEaling Care Homes Hospitals HICL UK Unitary ChargeEaling Schools, UK education HICL UK Unitary ChargeEast Down Colleges education BBGI UK Unitary ChargeEcole Centrale Supelec PPPProject, France

education HICL France Unitary Charge

Edinburgh Schools education JLIF, HICL UK Unitary ChargeELLAS Hospitals INPP UK Unitary ChargeELLAS 2 Hospitals INPP UK Unitary ChargeELLAS 3 Hospitals INPP UK Unitary ChargeELLAS 4 Hospitals INPP UK Unitary ChargeEnfield Schools education JLIF UK Unitary ChargeEnfield Street Lighting Gov Services JLIF UK Unitary ChargeEssex PFI SPV 1 education INPP UK Unitary ChargeEssex PFI SPV 2 education INPP UK Unitary ChargeExeter Crown Court, UK Gov Services HICL UK Unitary ChargeFalkirk NPD Schools education HICL UK Unitary ChargeFife Schools 2 PPP education HICL UK Unitary ChargeFife Schools, UK education HICL UK Unitary ChargeForth Valley Royal Hospital Hospitals JLIF UK Unitary ChargeGlasgow Hospital Hospitals HICL UK Unitary ChargeGlasgow Schools education JLIF UK Unitary ChargeGloucester Fire and Rescue, UK Gov Services HICL UK Unitary ChargeGloucester Royal Hospital Hospitals BBGI UK Unitary ChargeGolden Ears Bridge Roads BBGI Canada Unitary ChargeGoscote Hospitals INPP UK Unitary ChargeGreater Manchester PoliceAuthority

Gov Services HICL, JlIF UK Unitary Charge

Groningen Tax Office Gov Services JLIF Netherlands Unitary ChargeHarrow NRC Hospitals INPP UK Unitary ChargeHaverstock School, UK education HICL UK Unitary ChargeHealth and Safety Executive(HSE) Merseyside Headquarters


Health and Safety Laboratory,UK


Helicopter Training Facility, UK education HICL UK Unitary ChargeHereford and Worcester Gov Services INPP UK Unitary ChargeHighland School, Enfield education JLIF, HICL UK Unitary ChargeHome Office Headquarters, UK Gov Services HICL UK Unitary ChargeIrish Grouped Schools education HICL Ireland Unitary ChargeIslington I Housing Gov Services JLIF, INPP UK Unitary ChargeIslington II Housing Gov Services JLIF, INPP UK Unitary ChargeKelowna and Vernon Hospitals Hospitals BBGI, JLIF Canada Unitary ChargeKent PFI SPV 1 education INPP, BBGI, HICL UK Unitary ChargeKicking Horse Canyon Roads BBGI, HICL Canada Unitary ChargeKingston Hospital Hospitals JLIF UK Unitary ChargeKirklees Social Housing Gov Services JLIF UK Unitary ChargeKromhout Barracks Gov Services JLIF Netherlands Unitary ChargeLambeth Street Lighting Gov Services JLIF UK Unitary ChargeLancashire PFI SPV 1 education INPP UK Unitary ChargeLancashire PFI SPV 2 education INPP UK Unitary ChargeLancashire PFI SPV 2A education INPP UK Unitary ChargeLancashire PFI SPV 3 education INPP UK Unitary ChargeLeeds Combined SecondarySchools

education JLIF UK Unitary Charge

Lewisham Hospital Hospitals HICL UK Unitary ChargeLewisham PFI SPV 1 education INPP UK Unitary ChargeLewisham PFI SPV 2 education INPP UK Unitary ChargeLewisham PFI SPV 3 education INPP UK Unitary ChargeLisburn College education BBGI UK Unitary ChargeLiverpool and Sefton Clinics Hospitals BBGI UK Unitary ChargeLiverpool Library Gov Services INPP UK Unitary ChargeLong Bay Hospitals INPP Australia Unitary Charge




8. Appendices

Table 17: List of project finance SPVs in the listed PFI portfolio (continued)

Project name Sector Investor Country Revenue SourceLUL Connect Roads JLIF UK Unitary ChargeLuton PFI SPV 1 education INPP UK Unitary ChargeM40 Motorway Roads JLIF UK Unitary ChargeM6/M74 Project Roads JLIF UK Unitary ChargeM80 DBFO Roads HICL,BBGI UK Unitary ChargeMaesteg education INPP UK Unitary ChargeManchester School education HICL UK Unitary ChargeManchester Street Lighting Gov Services JLIF UK Unitary ChargeMedway LIFT Hospitals HICL UK Unitary ChargeMedway Police Gov Services HICL UK Unitary ChargeMersey Care Mental HealthHospital

Hospitals BBGI UK Unitary Charge

Metropolitan Police SpecialistTraining Centre

Gov Services HICL, JLIF UK Unitary Charge

Miles Platting Social Housing,UK


Ministry of Defence MainBuilding


Moray Schools education INPP UK Unitary ChargeN17/18 Road, Ireland Roads HICL Ireland Unitary ChargeNewcastle Hospital Hospitals JLIF UK Unitary ChargeNewcastle Libraries education HICL UK Unitary ChargeNewham Hospital Hospitals JLIF UK Unitary ChargeNewham Schools education JLIF, INPP UK Unitary ChargeNewport Schools education HICL UK Unitary ChargeNewton Abbot Hospital Hospitals HICL UK Unitary ChargeNorfolk Gov Services INPP UK Unitary ChargeNorth East Fire and Rescue Gov Services JLIF UK Unitary ChargeNorth Staffordshire Hospital Hospitals JLIF UK Unitary ChargeNorth Swindon Schools education JLIF UK Unitary ChargeNorth Tyneside Schools, UK education HICL UK Unitary ChargeNorth Wales Police Authority Gov Services INPP UK Unitary ChargeNorthampton Mental Health Hospitals JLIF UK Unitary ChargeNorthampton Schools education INPP UK Unitary ChargeNortheast Stoney Trail Roads BBGI Canada Unitary ChargeNorthwest Anthony HendayRing Road P3

Roads HICL, BBGI Canada Unitary Charge

Northwood MoD HQ Gov Services HICL UK Unitary ChargeNorwich Area Schools PFIProject


Nottingham PFI SPV 1 education INPP UK Unitary ChargeNottingham PFI SPV 2 education INPP UK Unitary ChargeNSW Schools education INPP Australia Unitary ChargeNuffield Hospital Hospitals HICL UK Unitary ChargeOldham Library education HICL UK Unitary ChargeOldham Secondary Schools PFIProject


Orange Hospital Hospitals INPP Australia Unitary ChargeOxford Churchill Oncology Hospitals HICL UK Unitary ChargeOxford Dunnock Way and EastOxford


Oxford John Radcliffe PFIHospital


Pembury Hospital Hospitals JLIF UK Unitary ChargePerth and Kinross Schools education HICL UK Unitary ChargePeterborough Hospital Hospitals JLIF UK Unitary ChargePeterborough Schools education JLIF UK Unitary ChargePforzheim Schools education INPP UK Unitary ChargePinderfields and PontefractHospitals, UK


QFTO - Barrow Renewables INPP UK unknownQFTO - Gunfleet Sands Renewables INPP UK unknownQFTO - Lincs Renewables INPP UK unknownQFTO - Ormonde Renewables INPP UK unknownQFTO -Robin Rigg Renewables INPP UK unknownQueen Alexandra Hospital,Portsmouth, UK


Queen Elizabeth Hospital Hospitals JLIF UK Unitary ChargeQueen’s (Romford) PFI Hospital Hospitals HICL UK Unitary ChargeRD901 Road, France Roads HICL France Unitary ChargeRealise Health (LIFT) Colchester Hospitals JLIF UK Unitary ChargeRedbridge and Waltham ForestLIFT


Redcar and Cleveland StreetLighting


Reliance Rail Rail Link INPP Australia unknownRenfrewshire Schools, UK education HICL UK Unitary ChargeRhonnda Cynon Taf Schools education HICL UK Unitary ChargeRoseberry Park Hospital Hospitals JLIF UK Unitary ChargeRoyal Children’s Hospital Hospitals INPP Australia Unitary ChargeRoyal School of MilitaryEngineering PPP Project, UK


Royal Women’s Hospital Hospitals BBGI Australia Unitary Charge




8. Appendices

Table 18: List of project finance SPVs in the listed PFI portfolio (end)

Project name Sector Investor Country Revenue SourceRoyal Women’s Hospital Hospitals BBGI Australia Unitary ChargeSalford and Wigan BSF Phase 1 education HICL, INPP UK Unitary ChargeSalford and Wigan BSF Phase 2 education HICL, INPP UK Unitary ChargeSalford Hospital, UK Hospitals HICL UK Unitary ChargeScottish Borders Schools education BBGI UK Unitary ChargeSheffield BSF, UK education HICL UK Unitary ChargeSheffield Hospital, UK Hospitals HICL UK Unitary ChargeSheffield Schools education HICL UK Unitary ChargeShowgrounds Gov Services INPP Australia Unitary ChargeSirhowy Way Roads JLIF UK Unitary ChargeSomerset PFI SPV 1 education INPP UK Unitary ChargeSouth Ayrshire Schools, UK education HICL UK Unitary ChargeSouth Bristol CommunityHospital


South East London Policestations


South Lanarkshire Schools education JLIF UK Unitary ChargeSouth West Hospital,Enniskillen


Southwark PFI SPV 1 education INPP UK Unitary ChargeSouthwark PFI SPV 2 education INPP UK Unitary ChargeSt Thomas More School education INPP UK Unitary ChargeStaffordshire LIFT Hospitals HICL UK Unitary ChargeSTaG PFI SPV 1 education INPP UK Unitary ChargeSTaG PFI SPV 2 education INPP UK Unitary ChargeStoke Mandeville Hospital, UK Hospitals HICL UK Unitary ChargeStoke on Trent and Stafford-shire Fire and Rescue Service

Gov Services BBGI UK Unitary Charge

Strathclyde Gov Services INPP UK Unitary ChargeSurrey Street Lighting Gov Services JLIF UK Unitary ChargeSussex Custodial Services, UK Gov Services HICL UK Unitary ChargeTameside General Hospital Hospitals HICL UK Unitary ChargeTameside PFI SPV 1 education INPP UK Unitary ChargeTameside PFI SPV 2 education INPP UK Unitary ChargeTor Bank School education BBGI UK Unitary ChargeTower Hamlets Schools education INPP UK Unitary ChargeTyne and Wear Fire Stations Gov Services HICL UK Unitary ChargeUniversity of Bourgogne,France


University of Sheffield Project,UK

Education HICL UK Unitary Charge

Unna Administrative Centre Gov Services BBGI Germany Unitary ChargeVancouver General Hospital Hospitals JLIF Canada Unitary ChargeVictoria Prisons Gov Services BBGI Australia Unitary ChargeWakefield Street Lighting Gov Services JLIF UK Unitary ChargeWalsall Street Lighting Gov Services JLIF UK Unitary ChargeWaltham Forest PFI SPV 1 education INPP UK Unitary ChargeWest Lothian Schools education HICL UK Unitary ChargeWest Middlesex Hospital, UK Hospitals HICL UK Unitary ChargeWillesden Hospital Hospitals HICL UK Unitary ChargeWolvehampton PFI SPV 1 education INPP UK Unitary ChargeWolverhampton and Walsall Hospitals INPP UK Unitary ChargeWolverhampton PFI SPV 2 education INPP UK Unitary ChargeWomen’s College Hospital Hospitals BBGI Canada Unitary ChargeWooldale Centre for Learning,UK


Zaanstad Penitentiary Insti-tution, The Netherlands

Gov Services HICL Netherlands Unitary Charge



References



References

l Amenc, N., F. Goltz, A. Lodh, and L. Martellini (2014). Towards smart equity factor

indices: Harvesting risk premia without taking unrewarded risks. Journal of PortfolioManagement 40(4).

l Amenc, N., F. Goltz, L. Martellini, and P. Retkowsky (2010). Efficient indexation: An

alternative to cap-weighted indices. EDHEC-Risk Institute Publications.

l Bender, J., R. Briand, F. Nielsen, and D. Stefek (2010, January). Portfolio of Risk Premia: A

New Approach to Diversification. The Journal of Portfolio Management 36(2), 17–25.

l Bianchi, R. J., G. Bornholt, M. E. Drew, and M. F. Howard (2014, May). Long-term U.S.

infrastructure returns and portfolio selection. Journal of Banking & Finance 42, 314–325.

l Bird, R., H. Liem, and S. Thorp (2013, April). The Tortoise and the Hare: Risk Premium

versus Alternative Asset Portfolios. Journal of Portfolio Management 39(3), 112–122.

l Bird, R., H. Liem, and S. Thorp (2014, May). Infrastructure: Real Assets and Real Returns.

European Financial Management 20(4), 802–824.

l Bitsch, F. (2012). Do Investors Value Cash Flow Stability of Listed Infrastructure Funds?

l Blanc-Brude, F. (2013). Towards efficient benchmarks for infrastructure equity invest-

ments. EDHEC-Risk Institute Publications, 88.

l Blanc-Brude, F., M. Hasan, and T. Whittaker (2016, March). Revenue and dividend

payout in privately held infrastructure investments. EDHEC Infrastructure InstitutePublications March.

l Chen, H.-C., S.-L. Chung, and K.-Y. Ho (2011, May). The diversification effects of

volatility-related assets. Journal of Banking & Finance 35(5), 1179–1189.

l Dechant, T. and K. Finkenzeller (2013). How much into infrastructure? Evidence from

dynamic asset allocation. Journal of Property Research 30(2), 103–127.

l Eun, C. S., W. Huang, and S. Lai (2008). International Diversification with Large- and

Small-Cap Stocks. Journal of Financial and Quantitative Analysis 43(02), 489–523.

l Fabozzi, F. J. and H. Markowitz (2011). The Theory and Practice of InvestmentManagement (2nd ed.). Hoboken, New Jersey: John Wiley & Sons.

l Faff, R. (2001, June). An Examination of the Fama and French Three-Factor Model Using

Commercially Available Factors. Australian Journal of Management 26(1), 1–17.



References

l Fama, E. F. and K. R. French (1992, June). The Cross-Section of Expected Stock Returns.

The Journal of Finance 47(2), 427–465.

l Fama, E. F. and K. R. French (1993, February). Common risk factors in the returns on

stocks and bonds. Journal of Financial Economics 33(1), 3–56.

l Finkenzeller, K., T. Dechant, and W. Schäfers (2010). Infrastructure: a new dimension

of real estate? An asset allocation analysis. Journal of Property Investment &Finance 28(4), 263–274.

l Gibbons, M. R., S. A. Ross, and J. Shanken (1989). A test of the efficiency of a given

portfolio. Econometrica 57(5), 1121–52.

l Huberman, G. and S. Kandel (1987). Mean-Variance Spanning. The Journal ofFinance 42(4), 873–888.

l Idzorek, T. and C. Armstrong (2009). Infrastructure and Strategic Asset Allocation: Is

Infrastructure an Asset Class? Technical report, Ibbotson Associates.

l Ilmanen, A. and J. Kizer (2012, April). The Death of Diversification Has Been Greatly

Exaggerated. The Journal of Portfolio Management 38(3), 15–27.

l Ince, O. S. and R. B. Porter (2006, December). Individual Equity Return Data From

Thomson Datastream: Handle with Care. Journal of Financial Research 29(4), 463–479.

l Jegadeesh, N. and S. Titman (1993, March). Returns to Buying Winners and Selling

Losers: Implications for Stock Market Efficiency. The Journal of Finance 48(1), 65–91.

l Kan, R. and G. Zhou (2012). Tests of Mean-Variance Spanning. Annals of Economicsand Finance 13, 139–187.

l Kroencke, T. A. and F. Schindler (2012, November). International diversification with

securitized real estate and the veiling glare from currency risk. Journal of InternationalMoney and Finance 31(7), 1851–1866.

l MSCI (2014). MSCIWorld Index. http://www.msci.com/resources/factsheets/index_fact_sheet/msci-

world-index.pdf.

l Newell, G., K. W. Chau, and S. K. Wong (2009). The significance and performance of

infrastructure in China. Journal of Property Investment and Finance 27(2), 180–202.

l Newell, G. and H. W. Peng (2007). The significance and performance of retail property

in Australia. Journal of Property Investment & Finance 25(2), 147–165.



References

l Newell, G. and H. W. Peng (2008). The role of US infrastructure in investment portfolios.

Journal of Real Estate Portfolio Management 14(1), 21–34.

l Oyedele, J., A. Adair, and S. McGreal (2014). Performance of global listed infrastructure

investment in a mixed asset portfolio. Journal of Property Research 31(1), 1–25.

l Peng, H. and G. Newell (2007). The significance of infrastructure in Australian

investment portfolios. Pacific Rim Property Research Journal 13(4), 423–450.

l Petrella, G. (2005). Are Euro Area Small Cap Stocks an Asset Class? Evidence from

Mean-Variance Spanning Tests. European Financial Management 11, 229–253.

l Rödel, M. and C. Rothballer (2012). Infrastructure as Hedge against Inflation—Factor

Fantasy? The Journal of Alternative Investments 15(1), 110–123.

l Rothballer, C. and C. Kaserer (2012a). Is Infrastructure Really Low Risk? An Empirical

Analysis of Listed Infrastructure Firms.

l Rothballer, C. and C. Kaserer (2012b). The Risk Profile of Infrastructure Investments:

Challenging Conventional Wisdom. The Journal of Structured Finance 18(2), 95–109.


About the EDHEC InfrastructureInstitute-Singapore




A Profound Knowledge GapEDHECinfra addresses theprofound knowledge gap

faced by infrastructureinvestors by collecting

and standardising privateinvestment and cash flow

data and runningstate-of-the-art asset

pricing and risk models tocreate the performance

benchmarks that areneeded for asset

allocation, prudentialregulation and the design

of new infrastructureinvestment solutions.

Institutional investors have set their sightson private investment in infrastructureequity and debt as a potential avenuetowards better diversification, improvedliability-hedging and reduced drawdownrisk.

Capturing these benefits, however, requiresanswering a number of difficult questions:

1. Risk-adjusted performance measuresare needed to inform strategic assetallocation decisions and monitoringperformance;

2. Duration and inflation hedgingproperties are required to understandthe liability-friendliness ofinfrastructure assets;

3. Extreme risk measures are in demandfrom prudential regulators amongstothers.

Today none of these metrics is documentedin a robust manner, if at all, for investorsin privately-held infrastructure equity ordebt. This has left investors frustrated byan apparent lack of adequate investmentsolutions in infrastructure. At the sametime, policy-makers have begun calling fora widespread effort to channel long-termsavings into capital projects that couldsupport long-term growth.

To fill this knowledge gap, EDHEC haslaunched a new research platform,EDHECinfra, to collect, standardise andproduce investment performance data forinfrastructure equity and debt investors.

Mission StatementOur objective is the creation a global repos-itory of financial knowledge and investmentbenchmarks about infrastructure equity and

debt investment, with a focus on deliv-ering useful applied research in finance forinvestors in infrastructure.

We aim to deliver the best availableestimates of financial performance andrisks of reference portfolios of privately-held infrastructure investments, and toprovide investors with important insightsabout their strategic asset allocationchoices to infrastructure, as well as supportthe adequate calibration of the relevantprudential frameworks.

We are developing unparalleled access tothe financial data of infrastructure projectsand firms, especially private data that iseither unavailable to market participantsor cumbersome and difficult to collect andaggregate.

We also bring advanced asset pricingand risk measurement technology designedto answer investors’ information needsabout long-term investment in privately-held infrastructure, from asset allocationto prudential regulation and performanceattribution and monitoring.

What We DoThe EDHECinfra team is focused on three keytasks:

1. Data collection and analysis: wecollect, clean and analyse the privateinfrastructure investment data of theproject’s data contributors as well asfrom other sources, and input it intoEDHECinfra’s unique database of infras-tructure equity and debt investmentsand cash flows. We also develop datacollection and reporting standards thatcan be used to make data collection




more efficient and reporting moretransparent.This database already covers 15 years ofdata and hundreds of investments and,as such, is already the largest dedicateddatabase of infrastructure investmentinformation available.

2. Cash flow and discount rate models:Using this extensive and growingdatabase, we implement and continueto develop the technology developedat EDHEC-Risk Institute to model thecash flow and discount rate dynamics ofprivate infrastructure equity and debtinvestments and derive a series of riskand performance measures that canactually help answer the questions thatmatter for investors.

3. Building reference portfolios ofinfrastructure investments: Usingthe performance results from ourasset pricing and risk models, we canreport the portfolio-level performanceof groups of infrastructure equity ordebt investments using categorisations(e.g. greenfield vs brownfield) that aremost relevant for investors’ investmentdecisions.

Partners of EDHECinfra

Monetary Authority of SingaporeIn October 2015, the Deputy Prime Ministerof Singapore, Tharman Shanmugaratnam,announced officially at the World BankInfrastructure Summit that EDHEC wouldwork in Singapore to create “usable bench-marks for infrastructure investors.”

The Monetary Authority of Singaporeis supporting the work of the EDHEC

Singapore Infrastructure InvestmentInstitute (EDHEC infra) with a five-yearresearch development grant.

Sponsored Research ChairsSince 2012, private sector sponsors havebeen supporting research on infrastructureinvestment at EDHEC with several researchChairs that are now under the EDHEC Infras-tructure Investment Institute:

1. The EDHEC/NATIXIS Research Chair onthe Investment and Governance Charac-teristics of Infrastructure Debt Instru-ments, 2012-2015

2. The EDHEC/Meridiam/Campbell LutyensResearch Chair on Infrastructure EquityInvestment Management and Bench-marking, 2013-2016

3. The EDHEC/NATIXIS Research Chairon Infrastructure Debt Benchmarking,2015-2018

4. The EDHEC/Long-Term InfrastructureInvestor Association Research Chair onInfrastructure Equity Benchmarking,2016-2019

5. The EDHEC/Global Infrastructure HubSurvey of Infrastructure Investors’Perceptions and Expectations, 2016

Partner OrganisationsAs well as our Research Chair Sponsors,numerous organisation have already recog-nised the value of this project and havejoined or are committed to join the datacollection effort. They include:

l The European Investment Bank;l The World Bank Group;l The European Bank for Reconstruction

and Development;l The members of the Long-Term Infras-

tructure Investor Association;




l Over 20 other North American, Europeanand Australasian investors and infras-tructure managers.

EDHECinfra is also :

l A member of the Advisory Council ofthe World Bank’s Global InfrastructureFacility

l An honorary member of the Long-termInfrastructure Investor Association

Origins and Recent AchievementsIn 2012, EDHEC-Risk Institute createda thematic research program on infras-tructure investment and established twoResearch Chairs dedicated to long-terminvestment in infrastructure equity anddebt, respectively, with the active supportof the private sector.

Since then, infrastructure investmentresearch at EDHEC has led to more than20 academic publications and as manytrade press articles, a book on infrastructureasset valuation, more than 30 industry andacademic presentations, more than 200mentions in the press and the creationof an executive course on infrastructureinvestment and benchmarking.

Testament to the quality of its contributionsto this debate, EDHEC infra’s research teamhas been regularly invited to contribute tohigh-level fora on the subject, including G20meetings.

Likewise, active contributions were made tothe regulatory debate, in particular directlysupporting the adaptation of the Solvency-2 framework to long-term investments ininfrastructure.

This work has contributed to growing thelimited stock of investment knowledge inthe infrastructure space.

Significant empirical findings alreadyinclude:

l The first empirical estimates ofconstruction risk for equity and debtinvestors in infrastructure projectfinance;

l The only empirical tests of the statis-tical determinants of credit spreads ininfrastructure debt since 2008, allowingcontrolling for the impact of marketliquidity and isolating underlying riskfactors;

l The first empirical evidence of thediversification benefits of investing ingreenfield and brownfield assets, drivenby the dynamic risk and correlationprofile of infrastructure investments overtheir lifecycle;

l The first empirical documentation of therelationship between debt service coverratios, distance to default and expecteddefault frequencies;

l The first measures of the impact ofembedded options in senior infras-tructure debt on expected recovery,extreme risk and duration measures;

l The first empirically documented studyof cash flow volatility and correlationsin underlying infrastructure investmentusing a large sample of collected datacovering the past fifteen years.

Key methodological advances include:




l A series of Bayesian approaches tomodelling cash flows in long-terminvestment projects including predictingthe trajectory of key cash flow ratios in amean/variance plane;

l The first fully-fledged structural creditrisk model of infrastructure projectfinance debt;

l A robust framework to extract the termstructure of expected returns (discountrates) in private infrastructure invest-ments using conditional volatility andinitial investment values to filter impliedrequired returns and their range atone point in time across heterogenousinvestors.

Recent contributions to the regulatorydebate include:

l A parsimonious data collection templateto develop a global database of infras-tructure project cash flows;

l Empirical contributions to adaptprudential regulation for long-terminvestors.


Infrastructure ResearchPublications at EDHEC



Infrastructure Research Publications atEDHEC

EDHEC Publications

l Blanc-Brude, F., T. Whittaker and M. Hasan. Cash Flow Dynamics of Private Infras-tructure Debt (March 2016).

l Blanc-Brude, F., T. Whittaker and M. Hasan. Revenues and Dividend Payouts inPrivately-Held Infrastructure Investments (March 2016).

l Blanc-Brude, F., and M. Hasan. The Valuation of Privately-Held InfrastructureEquity Investments (January 2015).

l Blanc-Brude, F., M. Hasan and O.R.H. Ismail. Performance and Valuation of PrivateInfrastructure Debt (July 2014).

l Blanc-Brude, F., Benchmarking Long-Term Investment in Infrastructure (June2014).

l Blanc-Brude, F., and D. Makovsek. How Much Construction Risk do Sponsors takein Project Finance. (August 2014).

l Blanc-Brude, F. and O.R.H. Ismail. Who is afraid of construction risk? (March 2013)

l Blanc-Brude, F. Towards efficient benchmarks for infrastructure equity invest-ments (January 2013).

l Blanc-Brude, F. Pension fund investment in social infrastructure (February 2012).

Books

l Blanc-Brude, F. and M. Hasan, Valuation and Financial Performance of Privately-Held Infrastructure Investments. London: PEI Media, Mar. 2015.

Peer-Reviewed Publications

l F. Blanc-Brude, S. Wilde, and T. Witthaker, “Looking for an infrastructure asset classDefinition and mean-variance spanning of listed infrastructure equity proxies”,2016 (forthcoming)

l Blanc-Brude, F., M. Hasan, and T. Witthaker, ”Benchmarking Infrastructure ProjectFinance - Objectives, Roadmap and Recent Progress”, Journal of AlternativeInvestments, 2016 (forthcoming)

l R. Bianchi, M. Drew, E. Roca and T. Whittaker, ”Risk factors in Australian bondreturns”, Accounting & Finance, 2015



Infrastructure Research Publications atEDHEC

l Blanc-Brude, F. “Long-term investment in infrastructure and the demand forbenchmarks,” JASSA The Finsia Journal of Applied Finance, vol. 3, pp. 57–65, 2014.

l Blanc-Brude, F. “Risk transfer, self-selection and ex post efficiency in publicprocurement: an example from UK primary and secondary school constructioncontracts,” Revue d’Economie Industrielle, vol. 141, no. 1st Quarter, pp. 149–180,2013.

l Blanc-Brude, F. , H. Goldsmith, and T. Valila, “A comparison of constructioncontract prices for traditionally procured roads and public–private partnerships,”Review of Industrial Organization, vol. 35, no. 1-2, pp. 19–40, 2009, ISSN: 0889-938X. DOI: 10.1007/s11151-009-9224-1.

l Blanc-Brude, F. , H. Goldsmith, and T. Valila, “Public-private partnerships in europe:an update,” EIB Economic & Financial Reports, p. 24, 2007.

l Blanc-Brude, F. and R. Strange, “How banks price loans to public-private partner-ships: evidence from the europeanmarkets,” Journal of Applied Corporate Finance,vol. 19, no. 4, pp. 94–106, 2007.

l Blanc-Brude, F. , H. Goldsmith, and T. Valila, “Ex ante construction costs in theeuropean road sector: a comparison of public-private partnerships and traditionalpublic procurement,” EIB Economic & Financial Reports, vol. 2006/1, 2006.

l O. Jensen and F. Blanc-Brude, “The handshake: why do governments and firmssign private sector participation deals? evidence from the water and sanitationsector in developing countries,” World Bank Working Papers, Wold Bank WorkingPaper Series, no. October 2005, p. 25, 2006.


For more information, please contact:Karen Sequeira on +65 6438 0030or e-mail: [email protected]

EDHEC Infrastructure Institute-SingaporeEDHEC Business School Asia-PacificOne George Street - #07-02Singapore 049145Tel.: +65 6438 0030

edhec.infrastructure.institute

Searching for a Listed Infrastructure Asset Class: Mean-variance ...

Documents